THE SOCIETY FOR PHOTONICSTechnologies (PhAST). PhAST is possibly the least prominent of these...

June 2008Vol. 22, No. 3www.i-LEOS.org IEEE

THE SOCIETY FOR PHOTONICS

NEWS

A NanophotonicInterconnect

for High-Performance

Many-CoreComputation

The National Ignition Facility(NIF) & Photon Science

Directorate at LawrenceLivermore National

Laboratory, USA

The National Ignition Facility(NIF) & Photon Science

Directorate at LawrenceLivermore National

Laboratory, USA

A NanophotonicInterconnect

for High-Performance

Many-CoreComputation

22leos03.qxd 6/16/08 1:15 PM Page cov1

22leos03.qxd 6/16/08 1:15 PM Page cov2

June 2008 IEEE LEOS NEWSLETTER 1

IEEE

THE SOCIETY FOR PHOTONICS

NEWS

Page 19, Figure 4

A simplified schematic example of an optical arbitration net-work used to provision four system resources. This scheme canrun in parallel to (and independently from) the network usedto transmit data between the cores. For example, each trans-mitter can determine whether a particular target receiver isavailable without communicating with the receiver. June 2008 Volume 22, Number 3

COLUMNS

Editor’s Column . . . . . . . . . . 2 President’s Column . . . . . . . . . . . 3

DEPARTMENTS

News . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24• Call for Nominations:

- 2009 John Tyndall Award- IEEE/LEOS 2009 Young Investigators Award- IEEE Medals and Recognitions

• Nomination Forms• IEEE Fellow First Woman Honored with John Fritz Medal• Memoriam: Prof. Pak Lim Chu, City University of Hong Kong

Careers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29• 2008 IEEE/LEOS Quantum Electronics Award Recipients:

Jeffrey H. Shapiro and Horace P. Yuen• 2008-2009 LEOS Distinguished Lecturers

Membership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35• Benefits of IEEE Senior Membership• New Senior Members• Chapter Highlight: LEOS Central New England Chapter

Conferences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38• Recognition at OFC 2008• Summer Topicals 2008• 2008 IEEE/LEOS International Conference on Optical MEMS &

Nanophotonics• 2008 IEEE/LEOS Avionics Fiber-Optics and Photonics Conference• ECOC 2008• LEOS Annual 2008• “Review of 2008 Winter Topical Meetings,” by Stefano Selleri,

University of Parma• LEOS Winter Topicals 2009• IPRM 2009

Publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46• Call for Papers:

- IEEE/Journal of Selected Topics in Quantum Electronics (JSTQE)

4

13

17

FEATURES

Research Highlights:“Photonics Work at Lawrence Livermore National Laboratory, USA,”

by Tony Ladran, LLNL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

“A Nanophotonic Interconnect for High-Performance Many-Core Computation,”by Ray Beausoleil, HP Laboratories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

22leos03.qxd 6/16/08 1:16 PM Page 1

Welcome to the June LEOS Newsletter! This month, wehave a feature article by Ray Beausoleil and colleagues atHP Laboratories describing a nanophotonic interconnectconcept that has great promise for improved computingperformance. We also have a summary of ongoing workin photonics at Lawrence Livermore National Laboratoryin the United States, including the National IgnitionFacility, home of the largest and highest-energy laser inthe world.

LEOS conferences provide excellent opportunities tointeract with fellow members and keep current with thelatest technology. Given the large number of LEOS meet-ings, most members cannot attend them all, so we arepresenting summaries and previews of meetings whenev-er possible. This month, we present highlights of theLEOS Winter Topical Meetings that took place inSorrento, Italy in January, and offer a preview of ECOCwhich will be held in Brussels, Belgium in September.

We are also having a contest! LEOS often uses figuresand images related to photonics in promoting the socie-ty. What better source of material than our members?Look for the contest announcement and details in thisissue and on the LEOS Web Portal.

As always, please feel free to send any comments andsuggestions to [email protected]. I would love tohear what you would like to see in future issues.

Krishnan Parameswaran

Editor’sColumnKRISHNAN PARAMESWARAN

PresidentJohn H. MarshIntense Photonics, Ltd.4 Stanley BoulevardHamilton International Tech ParkBlantyre, Glasgow G72 0BN, Scotland UKTel: +44 1698 772 037Fax: +44 1698 827 262Email: [email protected]

Secretary-TreasurerFilbert BartoliLehigh University19 West Memorial DrivePackard Lab 302Bethlehem, PA 18015Tel: +1 610 758 4069Fax: +1 610 758 6279Email: [email protected];[email protected]

Past-PresidentAlan WillnerUniversity of Southern CaliforniaDept. of EE-Systems/ Rm EEB 538Los Angeles, CA 90089-2565Tel: +1 213 740 4664Fax: +1 213 740 8729Email: [email protected]

Executive DirectorRichard LinkeIEEE/LEOS445 Hoes LanePiscataway, NJ 08855-1331Tel: +1 732 562 3891Fax: +1 732 562 8434Email: [email protected]

Board of GovernorsM. Amann H. KuwaharaK. Choquette C. MenoniC. Gmachl J. MeyerK. Hotate D. PlantJ. Jackel A. SeedsT. Koonen P. Winzer

Vice PresidentsConferences – E. GolovchenkoFinance & Administration – S. NewtonMembership & Regional Activities

- A. HelmyPublications – C. MenoniTechnical Affairs – A. Seeds

Newsletter Staff

Executive EditorKrishnan R. Parameswaran Physical Sciences Inc.20 New England Business CenterAndover, MA 01810Tel: +1 978 738 8187Email: [email protected]

Associate Editor of Asia & PacificHon TsangDept. of Electronic EngineeringThe Chinese University of Hong KongShatin, Hong KongTel: +852 260 98254Fax: +852 260 35558Email: [email protected]

Associate Editor of CanadaLawrence R. ChenDepartment of Electrical & Computer EngineeringMcConnell Engineering Building,Rm 633McGill University3480 University St.Montreal, QuebecCanada H3A-2A7Tel: +514 398 1879Fax: 514-398-3127Email: [email protected]

Associate Editor of Europe/MidEast/AfricaKevin A. WilliamsEindhoven University of TechnologyInter-University Research Institute COBRA on Communication TechnologyDepartment of Electrical EngineeringPO Box 5135600 MB Eindhoven, The NetherlandsEmail: [email protected]

Staff EditorGiselle BlandinIEEE/LEOS445 Hoes LanePiscataway, NJ 08855-1331Tel: +1 732 981 3405Fax: +1 732 562 8434Email: [email protected]

IEEE Lasers and Electro-Optics Society

LEOS Newsletter is published bimonthly by the Lasers and Electro-Optics Society of the Institute of Electrical and Electronics Engineers,Inc., Corporate Office: 3 Park Avenue, 17th Floor, New York, NY10017-2394. Printed in the USA. One dollar per member per year isincluded in the Society fee for each member of the Lasers andElectro-Optics Society. Periodicals postage paid at New York, NYand at additional mailing offices. Postmaster: Send addresschanges to LEOS Newsletter, IEEE, 445 Hoes Lane, Piscataway, NJ08854.

Copyright © 2008 by IEEE: Permission to copy without fee all or partof any material without a copyright notice is granted provided thatthe copies are not made or distributed for direct commercialadvantage, and the title of the publication and its date appear oneach copy. To copy material with a copyright notice requires spe-cific permission. Please direct all inquiries or requests to IEEECopyrights Office.

22leos03.qxd 6/16/08 1:16 PM Page 2


President’sColumnJOHN H. MARSH

I am writing this column shortly before the Conference on Laser andElectro-Optics (CLEO) held at the start of May in San Jose. Thisvenue is a new venue for CLEO, located in the heart of SiliconValley, and it will be interesting to see what impact the location willhave on the meeting. The early indicators are auspicious: the num-ber of submitted papers has increased and, in order to accommodatethe best of them, the length of the conference day has been extend-ed. Furthermore, the number of technical pre-registrations was sig-nificantly higher than at either the 2007 (Baltimore) and 2006(Long Beach) meetings.

In reality three conferences are co-located in San Jose: CLEOitself, the conference on Quantum Electronics and Laser Science(QELS) and the conference on Photonic Applications Systems andTechnologies (PhAST). PhAST is possibly the least prominent ofthese events, but is the conference experiencing the biggest changes.

Historically, PhAST has been well attended, but principallyby CLEO/QELS attendees. One of the aims of PhAST is to havea broad appeal and attract attendees in its own right. I have hadthe privilege of being one of this year’s PhAST co-chairs, alongwith David Roh, Andrew Masters, and David Huff. As in previ-ous years there are three parallel sessions, but this year one ofthem has been organized entirely by theOptoelectronics Industry DevelopmentAssociation (OIDA). As a result two of theparallel sessions address product develop-ment while the OIDA session addressesbusiness development.

The Product Development Sessions are• Lasers and LED Displays• High-Power Semiconductor Lasers• Trends in High-Power Diode Lasers • Lasers in Manufacturing• Laser Applications in the Photovoltaic

Market

The Business Development Sessions are• Organic LED Technology for Lighting• Business Growth for OLED Lighting• Organic LEDs for Low Power Displays• Organic Solar Cells• Inorganic Solar Technology and

Economics• New Solar Technologies for Grid

ParityThere is clearly some overlap, and hence syn-ergy, between the two groups of topics, par-ticularly in LEDs and photovoltaics, and it istherefore possible for attendees to crossbetween highly technical talks and businesssessions addressing the same topics.

In my April column, I described how LEOS is developinga long term strategic plan, and that a Strategic PlanningWorkshop would be taking place at OFC. I view develop-ment and implementation of this plan as a key goal of myPresidency. I also believe LEOS can and should build activi-ty in more application areas. As I write this column, theStrategic Plan has just been approved by the Board ofGovernors meeting at CLEO.

The goal for Technical Affairs is as follows‘Photonics practitioners will consider that LEOS activities reflect the fullscope of science, technology, and applications of photonics’

with the following objectives1.Improve response time for identifying emerging photonic

technologies.2.Increase scope of LEOS topics into underrepresented but

appropriate technical areas.3.Increase the activities targeted at the photonics markets.4. Increase cooperation with other disciplines that would bene-fit from photonics science and technology.

(continued on page 27)

22leos03.qxd 6/16/08 1:16 PM Page 3

4 IEEE LEOS NEWSLETTER June 2008

The National Ignition Facility (NIF) & PhotonScience Directorate at Lawrence Livermore NationalLaboratory comprises five programs, each focused on oneor more aspects of NIF's missions. They are:

National Ignition FacilityThe National Ignition Facility Project isresponsible for the construction and opera-tion of NIF, a 192-beam experimental laserfacility. This unique facility is the world's

largest and highest-energy laser, capable of creating temper-atures and pressures similar to those that exist only in thecores of stars and giant planets and inside nuclear weapons.

National Ignition CampaignThe National Ignition Campaign (NIC)encompasses all of the experiments, hard-ware and infrastructure needed to carry out

the initial ignition experiments on NIF beginning in2010 and to continue research on ignition in the follow-ing years. NIC is a key element of the National NuclearSecurity Administration.

Photon Science & ApplicationsThe Photon Science & Applications (PS&A)Program provides advanced solid-state laser and

optics technologies to the laboratory, government, and industry forimportant national needs. The primary activities of PS&A inrecent years have been: (1) to complete the laser technology ddevelopment and laser component testing for the U.S.Department of Energy's NIF project, (2) to develop advancedsolid-state laser systems and optical components for theDepartment of Defense and Department of Energy and (3) toaddress the needs of other government agencies and U.S. industry.

Inertial Fusion EnergyThe Inertial Fusion Energy Program isexploring a variety of approaches to usinginertial confinement fusion, NIF's coretechnology to achieve energy gain and help

lay the groundwork for the eventual use of fusion energyas a clean, safe, virtually limitless source of electricity.

Science at the ExtremesNIF will become a premier internationalcenter for experimental science early in thenext decade. The extreme temperaturesand pressures that will be created inside

the NIF target chamber will enable scientists from aroundthe world to conduct unprecedented experiments in highenergy density science and to gain new insights into suchmysterious astrophysical phenomena as supernovae, giantplanets, and black holes.

NIF: The ‘Crown Joule’ of Laser ScienceThe National Ignition Facility (NIF) is the world'slargest laser. See Figure 1. NIF's 192 intense laser beamswill deliver to its target more than 60 times the energy ofany previous laser system. When all 192 beams are oper-

Photonics Work at Lawrence Livermore National Laboratory

Research Highlights

THIS WORK PERFORMED UNDER THE AUSPICES OF THE U.S.DEPARTMENT OF ENERGY BY LAWRENCE LIVERMORE NATIONAL

LABORATORY UNDER CONTRACT DE-AC52-07NA27344

Figure 1. Three football fields could fit inside the NIF Laser andTarget Area Building.

Tony Ladran, Deputy of Operations, Photon Science and Applications Program,Lawrence Livermore National Laboratory

22leos03.qxd 6/16/08 1:16 PM Page 4

The language of technical computing.

Over one million people aroundthe world speak MATLAB.Engineers and scientists in everyfield from aerospace and semiconductors to biotech,financial services, and earth andocean sciences use it to expresstheir ideas.Do you speak MATLAB?

Parlez-vous MATLAB?

Quantitative high-throughput gene expressionimaging using data from FlyEx Database.

This example available atmathworks.com/ltc

Image from

FlyEx Database. U

sed by permission. http://flyex.am

s.sunysb.edu/flyex and http://urchin.spbcas.ru/flyex

®

©2007 The M

athWorks, Inc.

Accelerating the pace of engineering and science

22leos03.qxd 6/16/08 1:16 PM Page 5

ational in 2009, NIF will direct nearly two million joulesof ultraviolet laser energy in billionth-of-a-second pulsesto the target chamber center.

When all that energy slams into millimeter-sized tar-gets, it can generate unprecedented temperatures andpressures in the target materials – temperatures of morethan 100 million degrees and pressures more than 100billion times Earth's atmosphere. These conditions aresimilar to those in the stars and the cores of giant planetsor in nuclear weapons; thus one of the NIF & PhotonScience Directorate's missions is to provide a betterunderstanding of the complex physics of nuclear weapons.Researchers can also explore basic science, such as astro-physical phenomena, materials science and nuclear sci-ence. NIF's other major mission is to provide scientistswith the physics understanding necessary to create fusionignition and energy gain for future energy production.

A Variety of ExperimentsNot all experiments on NIF need to produce fusion igni-tion. Researchers are planning many other types of exper-iments that will take advantage of NIF's tremendous ener-gy and flexible geometry in non-ignition shots. Non-igni-tion experiments will use a variety of targets to derive abetter understanding of material properties under extremeconditions. These targets can be as simple as flat foils orconsiderably more complex. By varying the shock strengthof the laser pulse, scientists can obtain equation-of-statedata that reveal how different materials perform underextreme conditions for stockpile stewardship and basic sci-ence. They also can examine hydrodynamics, which is thebehavior of fluids of unequal density as they mix.

NIF experiments also will use some of the beams toilluminate "backlighter" targets to generate an X-rayflash. This allows detailed X-ray photographs, or radi-ographs, of the interiors of targets as the experimentsprogress. In addition, moving pictures of targets taken atone billion frames a second are possible using sophisticat-ed cameras mounted on the target chamber. These diag-

nostics can freeze the motion of extremely hot, highlydynamic materials to see inside and understand the phys-ical processes taking place As construction of the 48"quads" of four beams each proceeded, many shots werealready being performed using the first quad of beams.Experiments beginning in the winter of 2007-2008 willtake advantage of additional quads as they come online.

New Technologies Make NIF PossibleAmplifying NIF's beams to record-shattering energies,keeping the highly energetic beams focused, maintainingcleanliness all along the beam's path, and successfullyoperating this enormously complex facility – all requiredNIF's designers to make major advances in existing lasertechnology as well as to develop entirely new technolo-gies. Innovations in the design, manufacture, and assem-bly of NIF's optics were especially critical.

The Seven Wonders of NIFWhile construction of the football-stadium-sizedNational Ignition Facility was a marvel of engineering,NIF is also a tour de force of science and technologydevelopment. To put NIF on the path to ignition experi-ments in 2010, scientists, engineers and technicians hadto overcome a daunting array of challenges.

Working closely with industrial partners, the NIFteam found solutions for NIF's optics in rapid-growthcrystals, continuous-pour glass, optical coatings and newfinishing techniques that can withstand NIF's extremelyhigh energies. The team also worked with companies todevelop pulsed-power electronics, innovative control sys-tems and advanced manufacturing capabilities. Seventechnological breakthroughs in particular were essentialfor NIF to succeed:

1.Faster, Less Expensive Laser Glass ProductionLaser glass is the heartof the NIF laser system;it's the material thatamplifies the laser lightto the very high ener-gies required for exper-iments. NIF's laserglass is a phosphateglass that contains achemical additive withatoms of neodymium.The NIF laser system uses about 3,070 large plates oflaser glass. Each glass plate is about three feet long andabout half as wide. If stacked end-to-end, the plateswould form a continuous ribbon of glass 1.5 miles long.To produce this glass quickly enough to meet construc-tion schedules, NIF uses a new production method devel-oped in partnership with two companies – HoyaCorporation, USA and Schott Glass Technologies, Inc. –that continuously melts and pours the glass. See Figure3. Once cooled, the glass is cut into pieces that are pol-ished to the demanding NIF specifications.


Figure 2. Technicians adjust the target positioner inside the NIFTarget Chamber.

Figure 3. NIF Glass laser slab inproduction

22leos03.qxd 6/16/08 1:16 PM Page 6

22leos03.qxd 6/16/08 1:16 PM Page 7


2. Large Aperture Optical SwitchesA key element of the amplifier sectionof NIF's laser beampath is an opticaldevice called a plasma electrode Pockelscell, or PEPC, that contains a plate ofpotassium dihydrogen phosphate(KDP). See Figure 4. This device, inconcert with a polarizer, acts as a switch– allowing laser beams into the ampli-fier and then rotating its polarization totrap the laser beams in the amplifiersection. A thin plasma electrode that istransparent to the laser wavelength allows a high electric field tobe placed on the crystal, which causes the polarization to rotate.The trapped laser beams can then increase their energy muchmore efficiently using multiple passes back and forth throughthe energized amplifier glass. After the laser beams make fourpasses through the amplifiers, the optical switch rotates theirpolarization back to its normal configuration, letting themspeed along their path to the target chamber.

3. Stable, High-Gain PreamplifiersNIF uses 48 preampli-fier modules, or PAMs,each of which provideslaser energy for fourNIF beams. The PAMreceives a very lowenergy (billionth of ajoule) pulse from themaster oscillator roomand amplifies the pulseby a factor of about amillion, to a millijoule. It then boosts the pulse onceagain to a maximum of about ten joules by passing thebeam four times through a flashlamp-pumped rod ampli-fier. To perform the range of experiments needed on NIF,the PAMs must perform three kinds of precision spatial,spectral and temporal shaping of the input laser beams.

4. Deformable MirrorsThe deformable mirroris an adaptive optic thatuses an array of actuatorsto bend its surface tocompensate for wave-front errors in the NIFlaser beams. There is onedeformable mirror foreach of NIF's 192beams. Advances inadaptive optics in theatomic vapor laser isotope separation (AVLIS) program atLawrence Livermore National Laboratory demonstratedthat a deformable mirror could meet the NIF performancerequirement at a feasible cost. Livermore researchers devel-oped a full-aperture (40-centimeter-square) deformablemirror that was installed on the Beamlet laser in early

1997. Prototype mirrors from two vendors were also testedin the late 1990s. The first of NIF's deformable mirrorswere fabricated, assembled and tested at the University ofRochester's Laboratory for Laser Energetics and installedand successfully used on NIF to correct wavefronts for thefirst beams sent to target chamber center.

5. Large, Rapid-Growth CrystalsNIF's KDP crystals servetwo functions: frequencyconversion and polariza-tion rotation. The devel-opment of the technologyto quickly grow high-quality crystals shown inthis photo, were a majorundertaking and is per-haps the most highlypublicized technologicalsuccess of the NIF project. NIF laser beams start out asinfrared light, but the interaction of the beams with thefusion target is much more favorable if the beams are ultravi-olet. Passing the laser beams through plates cut from largeKDP crystals converts the frequency of their light to ultravi-olet before they strike the target. The rapid-growth processfor KDP, developed to keep up with NIF's aggressive con-struction schedule, is amazingly effective: Crystals that wouldhave taken up to two years to grow by traditional techniquesnow take only two months. In addition, the size of the rapid-growth crystals is large enough that more plates can be cutfrom each crystal, so a smaller number of crystals can provideNIF with the same amount of KDP.

6. Target FabricationTo meet the needs of NIFexperiments, NIF's mil-limeter-sized targetsmust be designed andfabricated to meet precisespecifications for density,concentricity and surfacesmoothness. When a newmaterial structure is needed, materials scientists create thenecessary raw materials. Fabrication engineers then deter-mine whether those materials – some never seen before – canbe machined and assembled. Manufacturing requirements forall NIF targets are extremely rigid. Components must bemachined to within an accuracy of one micrometer, or one-millionth of a meter. In addition, the extreme temperaturesand pressures the targets will encounter during experimentsmake the results highly susceptible to imperfections in fabri-cation. Thus, the margin of error for target assembly, whichvaries by component, is strict. Throughout the designprocess, engineers inspect the target materials and compo-nents using nondestructive characterization methods toensure that target specifications are met and that all compo-nents are free of defects. Together, this multidisciplinary teamtakes an experimental target from concept to reality.

Figure 5. 10 Joule PreAmpliermodule

Figure 6. 40 cm x 40 cm full aperturedeformable mirror

Figure 7. NIF rapid-growth KDPcrystal

Figure 8. Target Assembly

Figure 4. PlasmaElectrode Pockels Cell

22leos03.qxd 6/16/08 1:16 PM Page 8

22leos03.qxd 6/16/08 1:16 PM Page 9


7. Integrated Computer Control SystemFulfilling NIF's promiserequires one of the mostsophisticated computercontrol systems in gov-ernment service or pri-vate industry. Every NIFexperimental shot requiresthe coordination of complex laser equipment. In theprocess, some 60,000 control points for electronic, highvoltage, optical and mechanical devices – such as motor-ized mirrors and lenses, energy and power sensors, videocameras, laser amplifiers, pulse power and diagnosticinstruments – must be monitored and controlled. SeeFigure 9. The precise orchestration of these parts byNIF's integrated computer control system will result inthe propagation of 192 separate nanosecond (billionth ofa second)-long bursts of light over a one-kilometer pathlength. The 192 separate beams must have optical path-lengths equal to within nine millimeters so that thepulses can arrive within 30 picoseconds (trillionths of asecond) of each other at the center of a target chamberten meters in diameter. Then they must strike within 50micrometers of their assigned spot on a target measur-ing less than one centimeter long – an accuracy compa-rable to throwing a pitch over the strike zone from 350miles away.

Developing the State-of-the-art for National SecurityThe Photon Science and Applications (PS&A) program pursuesnational security missions by developing state-of-the-art opticsand laser technology. Through research and development,PS&A is developing technologies that advance the frontiers ofdiode-pumped laser technology, high-peak power lasers andscience, generation and application of bright, high-energy pho-tons and fabrication of precision, meter-scale optics.

Tailored-Aperture Ceramic LaserTo improve the run time and beam quality of large, high-average-power (HAP), diode-pumped solid-sate laser(DPSSL) systems, PS&A has developed a new slab laser tech-nology called the tailored-aperture ceramic laser (TACL). Ahigher-performance descendent of the solid-state heat-capac-ity The TACLLaser (SSHCL) system shown, TACL in Figure10 uses composite ceramic Nd3+:YAG/Sm3+:YAG slabsthat are edge pumped.

By smoothing the high spatial frequency pump-deposition rip-ples that result from face pumping, edge pumping can improve sys-tem beam quality and runtime considerably. New diode-pumparrays are also employed in TACL that use high-performancemicrolens conditioning and microchannel cooling. The microchan-nel cooling of the arrays allows the diode packages to be run with ahigh duty cycle – even continuously – while generating high aver-age optical pump power. As a result, TACL designs can use as lit-

x (−1)

Figure 10. Edge pumping is made possible by use of a ceramic Sm3+:YAG frame that is integral to the Nd3+:YAG portion of the slab.Sm3+:YAG absorbs at the laser wavelength, but is transparent at the pump wavelength. The Sm3+:YAG frame suppresses amplified sponta-neous emission without affecting passage of the pump light. This recent development proceeds from the concerted work of LLNL and theKonoshima Chemical Company and promises to revolutionize the design of large HAP laser systems by significantly improving beam quality.

Figure 9. NIF Control Room

22leos03.qxd 6/16/08 1:16 PM Page 10


tle as one-tenth of the diode arrays employed by the SSHCL systemwhile achieving the same power output capability.

Solid-State Heat-Capacity LaserPS&A is developing a high-average-power (100-kW-class),diode-pumped, solid-state heat-capacity laser (SSHCL) suitable foruse in military weapons. A mobile, compact, and lightweightSSHCL laser system capable of being deployed on a variety of plat-forms is also under development. Potential military applications ofsuch a system include the targeting and destruction of short-rangerockets, guided missiles, artillery and mortar fire, unmanned aeri-al vehicles and improvised explosive devices, or IEDs.

In 2006, PS&A achieved a major accomplishment when theSSHCL produced 67 kilowatts of power – a 50 percent increase inthe world-record-setting power level achieved the previous year.See Figure 11. This class of power demonstrates that tabletop-sizedsolid-state lasers have come of age and can fulfill the performancerequirements for their use in tactical weapon applications.

Additionally, improvements to the SSHCL's laser optics, bothin material selection and geometric architecture, have greatlyenhanced temperature profile uniformity throughout the lasingcavity, yielding a beam quality two times the diffraction limit forfive seconds of run time in PS&A's unstable resonator. See Figure12. Beam quality control is integral to PS&A's laser systemdevelopment, and results like these portend the use of directed-energy weapons on the battlefield where they can effect "speed-of-light" engagement in compact, mobile packages.

The extremely high fracture toughness of ceramicYAG:Nd3+, along with its facility for use in composinglarger slabs, makes it the ideal material for laser gain media.

Laser-Material InteractionPS&A's SSHCL test bed has afforded execution of extensivelaser-material interaction experiments using a wide varietyof materials such as steels, aluminums, titaniums andorganic composites. Varying the spot size of the laser beamfrom a few millimeters to a 162-centimeter locus while sus-taining an air flow rate of 100 meters per second at thepoint of laser-material interaction loosely simulates a targetflying in the atmosphere. See Figure 13. All tests were con-ducted at the nominal 25 kilowatts of average laser power.

ReducedAperture

Three0

20

40

Out

put P

ower

(kW

)

60

80

Three Four

Number of Slabs

Five

CalculatedMeasured @10%Duty CycleMeasured @20%Duty Cycle

Figure 11. The 67 kilowatts of average power was achieved usingfive ceramic neodymium-doped yttrium aluminum garnet(YAG:Nd3+) laser-gain media slabs; model calculations were vali-dated by experimental results as depicted in the graph.

22leos03.qxd 6/16/08 1:16 PM Page 11


These test results reveal the functioning of several mech-anisms when the laser is used to destroy a target: thermalheating of the high explosive to ignition; increased materi-al removal due to combustion of certain materials from the

increased oxygen during air flight; and aerodynamic forcesthat literally rip the membrane from the annealed andweakened target, making it aerodynamically unstable.

Potential PlatformsThe development of a prototype plat-form using SSHCL TACL technologysuitable for testing at a proving groundor test range is the next step in the evo-lution of directed-energy weapons. SeeFigure 14.

The transition from a laser-technol-ogy demonstration device in the labo-ratory to a fully-operational, directed-energy weapon capable of engaging livetargets under battlefield conditions is akey PS&A objective.

LLNL's work on the SSHCL heat capac-ity concept has set the stage for a new gen-eration of ceramic lasers and high-powerlaser architectures which will be capable ofrunning continuously at high efficiencyand with exceptional beam quality.

Diode-Pumped Alkali Laser:A New CombinationSince the advent of lasers more thanfour decades ago, solid-state and gaslasers have followed largely divergent

development paths. Gas lasers are based primarily ondirect electrical discharge for pumping (energizing),while solid-state lasers are pumped by flashlamps andsemiconductor diode laser arrays.

The alkali-vapor laser's intrinsically high efficiencyand its compatibility with today's commercially availablediode arrays enable fast-track development paths to tacti-cal systems, with mass-to-power ratios that far exceedwhat is possible with today's other laser systems.

Building on alkali-vapor laser research done by Z.Konefal, PS&A's Directed Energy Systems andTechnology (DEST) program element recently developeda new class of laser that combines features of both gas andsolid-state lasers, based on diode excitation of atomicalkali vapors. The defining features of the diode-pumpedalkali laser (DPAL) are its ability to be incoherentlypumped and its compatibility with diode arrays havingseveral-nanometer-wide spectral emissions. These charac-teristics distinguish DPALs from previous demonstrationsof alkali-based lasers that used narrow-band, coherentpumping to demonstrate lasing.

DEST's extensive laser modeling capability, anchored toexperimental laboratory demonstrations, supports extremepower scaling with good efficiency and beam quality.

DEST is the world leader in the development of thisnew class of laser; the first demonstration took place atLawrence Livermore National Laboratory in 2002, seeFigure 15, and it has been followed by many otherdemonstrations and developments.

Figure 13. SSHCL experimental test setup, using 25-kilowatt-average laser power, 132-cm laser beam spot size, and 100-meter-per-second air flow.

Figure 14. Concept of a 100-kW, movable, fully self-containedTACL/SSHCL ready for live target testing.

BK7 Window, No Diffusers FS Window, with Diffusers, Run #2BK7 Window, No Diffusers

00

1

2

3

4

5

6

7

8

200 400

Shot Number (200 Shots = 1 Second)

“Tim

es D

iffra

ctio

on L

imite

d” B

eam

Qua

lity

600 800 1000

Beam Quality

FS Window, with Diffusers, Run #2

Fused Silica Window, No Diffusers FS Window, with Diffusers, Run #3FS Window, with Diffusers, Run #1

No Diffusers, Power ~12 kW, with Diffusers, Power ! 10 kW, Unstable Resonator

FS Window, with Diffusers, Run #4

Fall 2005

Fall 2005

Spring 2006

Figure 12. A beam quality of two times the diffraction limit – for five seconds – wasachieved on the SSHCL testbed.

22leos03.qxd 6/16/08 1:16 PM Page 12


Fusion Energy Systems and ScienceFusion is the process by which the sun and stars produceenergy, and it is also responsible for much of the power ofthermonuclear weapons. It also has the potential to be asource of unlimited, environmentally friendly energy forhumankind.

The National Ignition Facility (NIF) is designed todemonstrate inertial confinement fusion, ICF, thatinvolves the rapid compression of small fuel capsules, asshown in figure 15, to reach densities and temperaturesgreater than those in the core of the sun; when a sufficientenergy density is reached, the fuel capsule ignites andthen burns while confined by its own inertia. NIF isexpected to demonstrate fusion ignition – the release ofmore energy via fusion burn than the laser energy used toinitiate ignition – early in the next decade, but will onlybe able to do so at a rate of one experiment every fewhours. For inertial fusion energy (IFE) production tobecome practical, targets will have to be ignited at a rateof several shots per second.

The mission of the Fusion Energy Systems and Science(FESS) program element of PS&A is to develop the laserdriver technology necessary to make IFE production prac-tical – including the development of low-cost, high-ener-gy, high-efficiency laser drivers and components for repe-tition rates of several times a second. Concurrent with thedevelopment of the National Ignition Facility is anotherambitious Livermore laser project named Mercury.

The Mercury laser project is an important part of theFESS mission. It is designed to produce 100-joule pulsesat a ten-per-second repetition rate with one kilowatt ofaverage power, using an architecture that could be scaledto IFE-relevant size. Mercury is a single-beam laser sys-tem, as shown in figure 16 that has developed capabilitiesthat will build on NIF's accomplishments. As currentlydesigned, NIF's 192 beams can fire simultaneously onlyonce every few hours. After each shot, the thousands ofoptics must be given a chance to cool down to ensure thatthey can operate correctly for the next shot.

Mercury has developed a method of continuously cool-ing the optics, while at the same time allowing the laserto fire rapidly over extended periods.

The current technology propels high-velocity helium gasacross the optics to keep them cool, while laser pulses passthrough the optics at a sustained rate of ten shots a second.

Unlike NIF, which uses seven-foot tall flashlamps toenergize the laser amplifiers, Mercury relies on diodelasers – similar to those in commercial read/write CDplayers – which give off one-third as much heat as flash-lamps. Mercury's beam is amplified as it passes throughslabs of specially grown ytterbium-strontium flouroap-atite crystals, as opposed to NIF's neodymium-dopedphosphate laser glass. More advanced amplifier media,such as transparent ceramics, are also being developed.

As of mid-2008, Mercury has been able to run contin-uously for several hours (300,000 shots), firing ten timesa second at more than 50 joules per shot, each shot last-ing just 15 nanoseconds (billionths of a second).

The project, which beganin 1996 and was initiallyfunded through LLNL'sLaboratory Directed Researchand Development (LDRD)office, has already been award-ed three R&D 100 Awards,most recently for developing aunique frequency conversioncrystal. Earlier awards were forthe original design of Mercury'sdiode array and for its Pockelscell, a unique light-switchingtechnology.

The long-term goal is alaser system capable produc-ing of NIF's energy output,with Mercury's ability torapidly fire shots, and ignite inertial fusion targets forelectrical power generation

High-Average-Power Laser-Diode ArraysFor 100-kW diode arrays to become a common reality, twoelements of the technology must be realized: high-perform-ance, reliable diode bars, and heat sinks that can sustain supe-rior thermal management and precision-diode bar mounting.

Figure 16. The Mercury laser system, a gas-cooled, diode-pumped, solid-state laser, operates in a facility about the size of a handball court.

Figure 15. NIF beam path intoa Hohlraum

22leos03.qxd 6/16/08 1:16 PM Page 13

Lawrence Livermore NationalLaboratory (LLNL) has devel-oped just such a package fordiode bars, using silicon – themainstay of the semiconduc-tor industry.

LLNL uses photolithog-raphy and etching tech-niques to produce tens ofthousands of 30-μm-widechannels in silicon sub-strates that carry coolingwater. The water flowingthrough these microchan-

nels significantly cools the laser-diode bars, which aremounted on the silicon less than 200 μm from the chan-nels. Combining ten diode bars onto a single heat sinkyields a ten-bar package (referred to as a "tile"), whichconstitutes the unit cell from which large, two-dimen-sional diode arrays can be built up through tiling.

The tiles used to make up these large arrays shown inFigure 18 are called silicon monolithic microchannels(SiMMs). Considerations that drove the SiMM packagedesign included ease of fabrication and the ability to con-struct large laser-diode arrays with output power capabil-ities of ten to 100 kW. Of paramount importance in the

design of the SiMM package was incorporation of thesame aggressive heat removal capability that character-ized LLNL's original, rack-and-stack silicon microchan-nel-cooled package. This was accomplished by placing themicrochannels into the silicon directly below the locationof the attached laser-diode bars. Like the rack-and-stacksilicon microchannel cooler, the SiMM design maintains avery tight thermal circuit, with just 177 μm of siliconseparating the heat-generating laser-diode bars and themicrochannel fins that define the cooling channels.

The SiMM laser-diode array is a packaging technology forproducing the smallest, most powerful and most inexpensivelaser-diode pumps ever. Each package of ten laser-diode barsintegrates the electrical, optical and hydraulic requirementsnecessary for high-average-power lasers. To date, arrays of up to100 kW have been fabricated at LLNL, and arrays of up to 100kW have been fabricated by SiMMtec, a commercial licensee ofthe technology.

With the relatively low cost of silicon, large arrays ofthese precision microchannels can be fabricated inexpen-sively using standard photolithography and etching tech-niques. Moreover, with silicon as the base material, indi-vidual diode bars can be precisely soldered to a package;each laser diode bar is connected to an LLNL-patentedmicrolens, giving the SiMM package its unsurpassedoptical brightness.


Silens Frame

Micro Lens

LightOutput

Metal

Notch

Water Inlet

CuRibbon

Diode

GlassManifold

Micro-Channels

Figure 18. A 41-kW array made up of 28 SiMM packages. The sketch illustrates a portion of a SiMM laser-diode package.

Figure 17. The first demonstra-tion of a resonance-transitionalkali laser using Rb vaporoccurred at LLNL in the winterof 2002.

22leos03.qxd 6/16/08 1:16 PM Page 14


AbstractSilicon nanophotonics holds the promise of revolutionizing com-puting by enabling parallel architectures that combine unprece-dented performance and ease of use with affordable power con-sumption. Here we describe the results of a detailed multiyeardesign study of dense wavelength division multiplexing (DWDM)on-chip and off-chip interconnects and the device technologies thatcould improve computing performance by a factor of 20 aboveindustry projections over the next decade.

1. IntroductionMoore’s Law is still a fundamental technology driver for the infor-mation technology industry: the ITRS Semiconductor roadmap [1]

shows, in the next decade and a half, a continued reduction in fea-ture sizes from 40 nm to the sub-10-nm regime. This growth incircuit density has brought us into the multi-core CPU era, and weare on the eve of a many-core (16 or more cores per socket) era. Astransistor density increases, the number of transistors comprising asingle computer core is not growing; rather, it has pulled back forreasons of power efficiency, allowing the number of cores packagedon the die to grow exponentially. In the many-core era, on-chip andoff-chip communication are the critical issues for sustaining per-formance growth for the demanding, data-intensive applications forwhich these many-core chips are intended. Computational band-width scales linearly with the exponentially growing number ofcores, but the rate at which data can be communicated across a chipusing top-level metal wires is increasing very slowly. The rate atwhich data can be communicated through pins at the chip edge isalso growing more slowly than the computational bandwidth, andthe energy cost of cross-chip and off-chip (from processor socket toDRAM) communication significantly limits the achievable band-width. As a result, the field of computer architecture is now in a cri-sis. It is not clear that many-core processors will find widespread

Research Highlights

1 INFORMATION AND QUANTUM SYSTEMS, HP LABORATORIES,1501 PAGE MILL RD., MS 1123, PALO ALTO, CA 94304–1123

2 EXASCALE COMPUTING, HP LABORATORIES, 1501 PAGE MILL RD., MS 1177, PALO ALTO, CA 94304–1177

3 EXASCALE COMPUTING, HP LABORATORIES, FILTON ROAD, STOKE GIFFORD, BRISTOL BS34 8QZ, UK

A Nanophotonic Interconnect for High-PerformanceMany-Core ComputationR. G. Beausoleil1, J. Ahn2, N. Binkert2, A. Davis2, D. Fattal1, M. Fiorentino1,N. P. Jouppi2, M. McLaren3, C. M. Santori1, R. S. Schreiber2, S. M. Spillane2,D. Vantrease2, and Q. Xu1

DIMM 1 DIMM 2 DIMM C

CPU

Memory Controller 2 Memory Controller CMemory Controller 1

Core1

Core2

Core3

Core4

Core5

Core6

Core7

Core8

CoreN−3

Cluster C’s L2 CacheCluster 2’s L2 CacheCluster 1’s L2 Cache

Optical Crossbar/Ring

N = 4C CoreN−2

CoreN−1

CoreN

Figure 1: Schematic of the basic architecture for a 256-core processor. The cores are divided into clusters that share an L2 cache and controlaccess to a particular unit of memory.

22leos03.qxd 6/16/08 1:16 PM Page 15


use: program performance may be consistently disappointing dueto limited communication bandwidths. In addition, exposing theselimited bandwidths to the programmer makes the parallel pro-gramming task far more difficult.

If we strive to continue delivering exponential performanceimprovements over a broad range of computational applicationsduring the next decade, then we are led inevitably to a symmet-ric architecture with the simplest possible crossbar interconnectthat allows the programmer to exploit parallelism at a high level.But this simplicity requires that we rely on an interconnect tech-nology that is no longer limited by the physics of copper wire.Here we argue that nanophotonics provides a feasible solutiontothis growing problem [2], and we present a simplified versionof a novel photonic interconnect architecture to be presented atISCA 2008 [3]. As we show in Section 6, the communicationbandwidth per unit of dissipated power provided by this on-chipoptical interconnect technology exceeds the maximum availablefrom purely electrical interconnects by a factor of 20 for applica-tions that heavily use the interconnect. Other proposals for on-chip nanophotonic interconnects have appeared recently, butmost simply replace the long “global” wires with a bus [4] or a

circuit-switched network [5]. These approaches do not efficientlyuse the advantages of nanophotonics and, as a consequence, havelimited bandwidth and large latency.

2. Model architectureIn the many-core era, the critical problems facing general purposecomputer design are heat generation and communication band-width. Architectural compromises forced by these issues makeprogramming difficult and impair application performance. Theprojected energy cost of cross-chip electrical communication (inthe year 2017) is expected to be 5.8 pJ/b; although an extremeRambus solution for off-chip signaling could achieve (in princi-ple) 2.2 pJ/b, current practice reaches about 20 pJ/b [6]. At thesecosts, all of the available power will be consumed moving dataunless a severe bandwidth choke is imposed. With optics, howev-er, over the next decade we believe that we can reduce the energyrequired for cross-chip communication to less than 0.5 pJ/band to0.1 pJ/b for off-chip communication. By exploiting the ability tomove data affordably, wherever it is needed, we can reduce bothpower and total cost while improving bandwidth, performance,and programmability.

Laser

Star Coupler 2Star Coupler 1

DetectorsDetectors

Data&

Control

Arbitration

Splitters

Splitters

Off-Chip

Broadcast

L1-I

Core 0Core 2

Through Silicon Via Array ModulatorsModulators

Modulators

Modulators

Modulators

Modulators

Modulators

ModulatorsModulators

Modulators

Modulators

4-WaveguideBundles

0123

N−1

N+1

N

616263

Detectors

64 VariableDrop Filters/

Detectors2 Fixed Drop

Filters

Detectors

L2Cache

Core 1

L1-I

L1-I L1-I

L1-D

L1-D

L1-D

L1-D

Pee

r X

-Bar

Con

nect

ion

NIMC

L1↔

L2 In

terf

ace

L1↔

L2 In

terf

ace

My

X-B

arC

onne

ctio

n

Directory

Hub

Figure 2: The nanophotonic interconnect die for a 256-core chip operating at 5 GHz. Subsystems illustrated here include: the 64 clustertiles on the processor and memory dies; 270 identical parallel ridge waveguides allocated to cache line transfers and control messages (blue),and additional waveguides providing off-chip I/O (green); an expanded view of a photonic data/control block, an arbitration block, abroadcast block, and an off-chip communication hub; and laser power distribution.

22leos03.qxd 6/16/08 1:16 PM Page 16


In Fig. 1, we show a block diagram of a many-core processorillustrating the basic idea behind the architecture considered here.The device is tiled with 64 identical four-core compute clusters,each of which has a memory controller that either interfaces tostacked memory or drives a photonic connection to off-chip mem-ory to provide bandwidth that scales with processor performance.We interconnect the clusters with a photonic crossbar, offeringenormous bandwidth, modest latency, and very low power con-sumption. This creates a symmetric architecture in which all mem-ory is close to all processors; the programmer expresses parallelismat a high level, and is not burdened by issues of locality, therebygreatly reducing the difficulty of parallel program development.Furthermore, we can provide bandwidth of one byte per double-precision floating-point operation (flop) all the way to DRAM,eliminating the need to exploit very complex software techniques(multilevel tiling, for example) to guarantee locality of reference.The bandwidth provided from DRAM is large enough to eliminatean entire layer of cache (L3), with significant savings in power andcost, while also reducing latency and hardware complexity.

Energy efficiency is the primary motivation for many-core, andsmall, power-efficient cores and caches achieve the best possibleperformance per unit energy. We don’t anticipate significantincreases in CPU clock rates over the next decade, so we assume forour model that the cores use a 5 GHz clock. If we also assume thatthe cores are dual-issue, in-order, and multithreaded, and that theyoffer SIMD instructions allowing 4 multiply-accumulate and 4-word-wide load/store operations, then the compute bandwidth ofthe device is 10 teraflops (i.e., double-pre-cision floating-point operations per second),requiring (at one byte per flop) 20 TB/s ofbidirectional interconnect bandwidth. Ourpower models predict that this processorwill dissipate approximately 200 W in thesilicon itself.

3. The nanophotonic crossbarDespite the remarkable progress maderecently in broadband Mach-Zehnder elec-trooptic modulators [7] and quantum-wellmaterials for electroabsorption modulators[8,9], we believe that — in the long term —the use of dense wavelength-division multi-plexing (DWDM) in optical interconnects isinevitable. In order to minimize the need forbuffering (and therefore eliminate the powerdissipated by serialization and deserializa-tion, or SERDES), we assume that we will beable to drive each physical channel at twicethe clock frequency (i.e., read or write twobits per clock to/from digital registers), or 10Gb/s. Therefore, we will need 16,000 physi-cal channels in our interconnect to supply therequired bidirectional bandwidth of 20 TB/s.However, in order to implement the crossbarwith the performance specifications (particu-larly low latency) required by our architec-ture, we will need to avoid switching alto-gether and overprovision the physical chan-

nels significantly. For example, in order to provide an all-to-all sin-gle-hop interconnect topology, we will require total direct connec-tions between the clusters. (In other words, at any given time, eachcluster may be sending data to only one of the other clusters, buteach cluster must always be connected physically to all 63 of theother clusters.) If we assume that we are using dies, and that eachwaveguide (or wire) occupies an area with dimensions 2 cm × 3 µmon average, we find that the interconnect will require over 200 lay-ers on the die, which is clearly infeasible, regardless of whether weare using wires, direct optical modulation, or optical coarse wave-length division multiplexing (CWDM). As we shall see next, usingDWDM in our design allows us to provide extraordinarily highinterconnectivity in only a single layer.

As illustrated to scale in Fig. 2, one implementation of thenanophotonic crossbar employs 64 wavelengths multiplexed over270 parallel waveguides, with 256 waveguides allocated to controland data, 2 to broadcast, and 12 to arbitration. Each cluster “lis-tens” to a single dedicated bundle of 4 waveguides, at all wave-lengths. These 256 individual bit-wide channels are grouped intological channels for data and control messages. The timing of theselogical channels is synchronous, each channel being used by onesender to send to its destination in any given 24 clock “epoch.” Anew, distributed, all-optical crossbar arbitration scheme describedin Section 5 allocates the logical channels to sending clusters, allow-ing one sending cluster to transmit to a given destination cluster ina given epoch. Once granted permission by the arbiter to transmit,the sending cluster modulates all wavelengths on the allocated

Corning Incorporated is a diversified technology company with a rich history spanning more than 150 years. Corning combines our expertise in specialty glass, ceramic materials, polymers and the manipulation of properties of light with strong process and manufacturing capabilities to develop, engineer and commercialize significant innovative products for the flat panel display, telecommunications, environmental and life sciences markets.

We are seeking an applied scientist to conduct optical physics research and to provide expertise in fundamental understanding of laser interactions with materials of fundamental importance to Corning, including glasses and ceramics. This position will also have the opportunity to perform exploratory research to extend our understanding of laser processing.

Responsibilities:

• Lead and/or support R&D teams to develop laser process technologies to enable next generation product or process features

• Direct special investigations of laser system capabilities related to

product or process performance including laser output and laser/material interaction

• Develop independent, exploratory research efforts to characterize and explore new laser/material interactions

Required Skills:

• Expertise in high power lasers and material interaction (e.g. drilling, ablation, etc.)

• Experimental experience with high power lasers and optical beam delivery systems

• Ability to develop and lead independent research effort to support laser processing initiatives

Education and Experience:

• Ph.D. in physics, electrical engineering, materials science or related field

• 5+ years experience with multiple types of laser systems (UV, CO2, diode and/or Nd:YAG) including beam delivery systems.

• 3+ years experience in field of laser processing of materials

Submit your application online at: http://www.corning.com/careers/index.aspx and search ref #160941

22leos03.qxd 6/16/08 1:16 PM Page 17


channel in order to send to the destination cluster. This on-chip,low-cost, high-bandwidth, low-power crossbar is able to handle asmany as 64 inputs and outputs simultaneously, and is a revolution-ary development that will be significant in many applications; itenables the kind of flattened, symmetric architecture we desire.

The same high bidirectional bandwidth to off-stack memory isa requirement for our target performance. Even with the number ofpins growing by 40 percent every generation, electrically-connect-ed memory will in 2017 provide only one fifth of a byte per flop tothe processor. To fully exploit the advantages of optics, we reorgan-ize memory into stacks with a photonic interface below the layersof DRAM, connected by fiber to the memory ports of the many-core processor stack. This approach provides adequate bandwidth ofabout one byte per flop (and eventually even more if necessary),solving the chip-edge bandwidth problem. It does so while savingconsiderable power (up to 200 W if communicating at 10 TB/selectrically at 2.2 pJ/b). Optical interfaces enable asingle DRAMchip to source an entire cache line, making better use of the largeinternal DRAM memory bandwidth and reducing power even fur-ther. This will provide the bandwidth and power benefits of aprocessor-in-memory (PIM) architecture while keeping the pro-grammability of a classical symmetric multiprocessor.

4. Nanophotonic componentsThis ultra-high-bandwidth DWDM network is enabled by a num-ber of CMOS-compatible nanophotonic devices that are (at most)only a few years beyond the current state of the art: 1. Low-loss silicon-on-insulator (SOI) waveguides have measured

losses as low as 0.2 dB/cm [10], and will not need muchimprovement. However, the commercially available SOI wafersused for this purpose today are custom-made to satisfy lightconfinement requirements by increasing the thickness of thetraditional oxide buffer layer, and are therefore relatively expen-sive. In addition, this thick oxide layer confines heat within thethin top silicon layer, which translates into a large temperaturebuild-up for sensitive resonant devices such as drop filters, mod-ulators, and ring lasers. Therefore, in the long term, it will beimportant to develop means for creating nanophotonic compo-nents using pure silicon wafers, which provide high thermalconductivity at low cost.

2. Resonant receiver-less Ge detectors. The smaller bandgap ofGe than Si (0.7 eV vs. 1.1 eV) allows detection of opticalsignals in the 1300 nm wavelength range, and with prop-erly designed material can likely extend to 1550 nm. Thepotential ability to fabricate Ge layers on Si offers the pos-sibility of integrating optical components with Si integrat-ed circuits for efficient electric-optical transduction. Ge canbe embedded into resonant detectors to allow a single wave-length in a waveguide to be detected. This also enables low-capacitance detectors that eliminate the need for power-hungry amplifiers and clock recovery to build a receiver-lessdetection scheme [11].

3. Resonant modulators. As described below, we propose to usering resonators that selectively modulate a single wavelengthon a given waveguide and can be moved to an “OFF” statewhere they are transparent to the data flux in the waveguide.The modulators work by changing the index of silicon ringsusing charge injection. Published results indicate that thetarget modulation rate of 10 Gb/s can be achieved with cur-rent technology [12]. These rings will need to be kept reso-nant with the chosen wavelength by thermal tuning of therefractive index.

4. Multiwavelength lasers with precisely controlled frequencyintervals are ideal for low-cost DWDM systems. If only one ofthe frequency channels is servo-locked to an on-chip standardcavity, then all of the other frequency modes will track the con-trolled mode. One of the possible approaches is the Fabry-Perotcomb laser based on quantum dots [13], which has already beenused to demonstrate a bit-error-rate of at 10 Gb/s over ten lon-gitudinal modes [14,15]. Another possible approach is themode-locked hybrid Si/III-V evanescent laser [16], which uses asilicon-waveguide laser cavity wafer-bonded to a III-V gainregion. In this case, any ambient temperature change in theenvironment will cause approximately the same refractive indexshift in the laser cavity and the silicon waveguides and res-onators that form the DWDM network. Our design studyshows that the laser need only provide 1–2 W of total opticalpower to supply a 20 TB/s network if the detector capacitanceis low enough that only 30,000 photons are needed to drive a 1V swing at the detector’s output terminal.

In our design of a single-layer DWDMnanophotonic interconnect, we have chosenthe silicon microring resonator as our founda-tional component because it has small size,high quality factor Q, transparency to off-res-onance light, and no intrinsic reflections.Using injected charge, the refractive index ofthe microring can be changed, shifting thefundamental frequency of the cavity eitherinto or out of resonance with an incident lightfield. The microring can act as an optical fil-ter [17,18], and it can be made into elec-trooptic modulators [12,19,20], lasers [21]and detectors when carrier injection, opticalgain, or optical absorption mechanisms areincorporated. In the past, the key characteris-tic of a silicon microring resonator yet to be

(a) (b)

Acc.V3.00 kV

Spot3.0

Magn65000x

DetTLD

WD5.3

Exp5 LSN#5 after etch

1 μm

Figure 3: (a) An SEM picture with 40°-tilted view of a microring resonator with a 1.5-µmradius coupled to a waveguide with an optimized (reduced) width. (b) A microscope pictureof cascaded microring resonators coupled to a U-shaped waveguide at the edge of the chip.

22leos03.qxd 6/16/08 1:16 PM Page 18


demonstrated experimentally is a radius approaching the minimumpossible value that allows an intrinsic Q of 20,000, which is about1.4 µm. As shown in Fig. 3(a), we have fabricated Si microringswith radii of 1.5 µm [22] with intrinsic Qs of 18,000 and effectivemode volumes around 1.0 µm. When coupled to an optimally-designed silicon strip waveguide that minimizes spurious lightscattering and increases the critical dimensions of the geometry(easing fabrication requirements), the coupled Q approaches thetheoretical maximum possible value for a ring of that size (9,000out of 10,000). In Fig. 3(b), we show cascaded silicon microring res-onators that can be used as a modulator or fil-ter bank in a nanophotonic network.

In the case where we use the microring asa modulator, a small size is critical for severalreasons. First, a smaller size means that moremodulators can be fit into a given area, there-fore providing higher integration density.Second and more importantly, the power con-sumption of the modulator, which is a keyperformance factor for electrooptical modula-tors, is directly proportional to the circumfer-ence and inversely proportional to the Q ofthe resonator. Reducing the size of the ringwithout sacrificing the Q is critical for low-power operation. Third, the total bandwidthof a microring-based DWDM modulationsystem [20] is limited by the free spectral

range (FSR) of the microring resonator, which is inversely propor-tional to the circumference of the ring. A smaller microring mod-ulator has a larger FSR, which can therefore accommodate morewavelength channels and have higher aggregate data bandwidth. Inour case, the choice of a 1.5 µm radius and the demonstration of thenear-maximum-possible coupled Q of 9,000 provides a FSR ofabout 8 THz, and a filter bandwidth of about 20 GHz, which isnearly ideal for our interconnect architecture.

One of the most important requirements that must be met bya nanophotonic interconnect is that its total thermal dissipation

EpochML Laser Component 1

Component 3

(a)

(c)

(b)

(d)

Com

pone

nt 4

Com

ponent 2

1

Component Frequency

2

3

4

1

Comp Bid Win

2

3

4

Epoch 1ML Laser Component 1

Component 3

Com

pone

nt 4

Com

ponent 2

1

2

3

4

1

2

3

4


Component 3

Com

pone

nt 4

Com

ponent 2

1

2

3

4

1

2

3

4

4

4

2

1


Component 3

Com

pone

nt 4

Com

ponent 2

1

2

3

4

1

2

3

4

4

3

4

1

Component Frequency

Component Frequency Component Frequency

Comp Bid Win Comp Bid Win

Comp Bid Win

Figure 4: A simplified schematic example of an optical arbitration network used to provision four system resources. This scheme can runin parallel to (and independently from) the network used to transmit data between the cores. For example, each transmitter can determinewhether a particular target receiver is available without communicating with the receiver.

4 16 64 64 4 4 4 16

0.64 2.56 10.24 40.96

1.28 5.12 20.48 81.921.8 4.3 8.9 27.5 177 105 54 42

1.28 5.12 20.48 81.923.4 18.0 38.4 118.4332 439 235 181

Technology Node (nm)

Clusters/chipCores/clusterComputational performance (Tflops)

On-chip interconnectBi-directional bandwidth (TB/s)Power (W)Energy/bit (fJ)

Off-chip interconnectBi-directional bandwidth (TB/s)Power (W)Energy/bit (fJ)

40 nm 28 nm 17 nm 14 nm

TABLE 1: Projected DWDM interconnect performance at various technology nodes. Thepower consumption of the interconnect includes lasers and modulators, as well as the powerneeded to keep resonant detectors and modulators locked.

22leos03.qxd 6/16/08 1:16 PM Page 19


remain below 25% that of the underlying silicon transistors, or lessthan 50 W for our target processor in 2017. As discussed above,we expect that the laser source will contribute about 5 W to thistotal, but the on- chip microrings will contribute a much largerquantity of heat. There are three possible modes for electricalpower dissipation in rings: fabrication error trimming, resonancefrequency biasing, and direct data modulation. Because of fabrica-tion imperfections, each ring will have a resonance frequency thatis slightly different from the design goal, and must be “trimmed”into the correct spectral location. We can rely on two schemes tofine-tune the rings: we can use carrier injection to blue-shift theresonance, and thermal heating to red-shift it. In the worst-casescenario we need 185 µW/nm to red-shift a 3-µm ring throughheating and 125 µW/nm to blue-shift it through current injec-tion. The approach that combines heating and current injection,however, is only viable if the critical dimension control of the fab-rication process is better than 1 nm. If this condition is not met,then thermal control alone needs to be used at the expense of amuch larger power consumption.

The electric current required to modulate a ring is given byC/pulse at 5 GHz, or 30 fJ at 1 V, assuming that we detune the ring40 GHz from resonance to obtain an extinction ratio of 10 dB orgreater. This corresponds to a raw dissipated power of 150 µW peronline ring. However, the modulator voltage driver circuit will nec-essarily dissipate electrical power, since the modulator acts as acapacitive load with a 10 µA leakage current in the “on” state, andhas a peak current of 1 mA during the transition. Recently, thisproblem has been solved on-die by manufacturers of packagedCMOS photonic devices using AD-DA conversion drivers, but thepower dissipated in these drivers has been much greater than thepower expended in the modulators themselves. Therefore, in thelong term we believe that it will be important to develop purelyanalog CMOS drivers to reduce the electronic overhead by a factorof 30–100 over the current state-of-the art. We believe that the effi-ciency of these drivers will scale only slowly with circuit featuresize, resulting in a modulation power below 0.5 mW at the 17-nmtechnology node.

Generally, in the reconfigurable network we will trim all rings(i.e., both modulators and detectors) away from resonance, andthen use current injection to bring the necessary rings online oncearbitration is complete. Given the silicon parameters mentionedabove, and that an active ring will be online during the entireepoch, the total power requirement per online ring is about 30µW, or about 0.1 mW including analog driver overhead at the 17-nm technology node. Therefore, in 2017 we expect the power dis-sipated by all on-chip rings to be approximately 40 W. We have

modeled the performance of the on-chip and off-chip interconnectsshown in Fig. 2 at several technology nodes in Table 1. The totalpower consumption is the sum of the on-chip and off-chip esti-mates, and includes all of the laser, modulation, and trimmingcontributions outlined above.

5. All-optical arbitrationOne of the most significant contributions to the interconnect laten-cy is the time required to determine the availability of systemresources, arbitrate collisions between requests, and then grantaccess to the resource requestors. For example, in an all-to-all mul-tihop switched interconnect architecture (e.g., a torus), electricalsignals representing requests must be sent to an arbitration proces-sor, and wait for the outcome of a computation and then receipt ofanother electrical signal before transmission can commence.However, a key advantage of our solution is that — at the cost ofoverprovisioning the nanophotonic components — the transmit-ters themselves can reconfigure the crossbar in a few clock cycles,thus avoiding the need for hand-shaking procedures that increaselatency. Nevertheless, we still require an arbitration system to han-dle collisions. The intrinsic parallelism of optical signals allows usto propose a novel, all-optical, low-latency arbitration system thatdoes not require digital electronic computation or communicationbetween the transmitter and receiver, using a protocol that can berun completely independently of the on-chip data network. Ouranalysis shows that this protocol provides nearly the best possiblethroughput under light and moderate loads, and about 90% of thebest possible throughput under heavy loads.

A schematic diagram of a simple version of optical arbitration isshown in Fig. 4 for the case where there are four system resources(e.g., L2 caches in cores or clusters of cores) to be allocated. A four-wavelength (e.g., mode-locked) laser provides optical power to eachcomponent in a single distribution waveguide, and each wave-length is dropped onto the arbitration waveguide at a specific loca-tion in the ring. For example, the “red” wavelength (which in factmay belong to a particular channel near 1310 nm) is alwaysdropped onto the arbitration waveguide by a microring resonatornear component 1 that is always tuned to that wavelength. At thebeginning of the first 24-clock-cycle “epoch,” each resource isassigned a unique wavelength using a predetermined algorithmknown to each component, and each component prepares a “bid”for a particular resource (in this case, the right to transmit data toanother component). In an electrical arbitration system, this bidwould be an electrical signal sent to a dedicated subprocessor, buthere the bid is made locally by tuning an adjacent drop filter to thewavelength assigned to the desired target resource. For example,

here component 1 bids for resource 4 by acti-vating (i.e., tuning into precise resonance) alocal microring resonator that is designed todrop the wavelength currently assigned toresource 4 (i.e., “blue” during this epoch) ontoan integrated photodetector. If the opticalpower sensed by the photodetector rises abovea designated threshold, then component 1 haswon the right to transmit to resource 4.However, in this epoch, component 2 alsobids for access to resource 4; since component2 is “upstream” from component 1, and the

BenchmarkOptical

PerfomanceElectrical

PerformanceScaled

Optical/Electrical

PTRANS (GB/S) 9102 459.0 22STREAM (GB/S) 10240 605.0 19GUPS 40 2.4 19DGEMM (Gflops) 37 37.0 1FFT (Gflops) 1734 879.0 2MPI (GB/s) 20 1.2 19

TABLE 2: HPCS benchmark performance for the proposed architecture at the 17-nmtechnology node.

22leos03.qxd 6/16/08 1:16 PM Page 20


“blue” wavelength is dropped onto the arbitration waveguide nearcomponent 3, it is component 2 that wins the arbitration round andaccess to resource 4. Since the optical power at the “blue” wave-length is removed from the waveguide by component 2, the “blue”photodetector sees a low optical intensity, and must wait until thenext epoch to try again to transmit data to resource 4. However,during the next epoch, the wavelengths representing the availableresources are reassigned, and now component 1 wins the right totransmit to resource 4 even though component 3 tries to bid for thesame resource. A more sophisticated token-ring optical arbitrationscheme [3] eliminates the need to synchronize execution duringepochs, and allows latency to be reduced even further.

6. System performance modelsWe have modeled the performance of the nanophotonic architec-ture described in the previous section (as well as an idealized elec-trical equivalent) for the HPC Challenge benchmarks [23], whichtypify high-performance data access patterns. In our model, we cal-culated performance limits due to CPU, interconnect, and memo-ry bandwidths. We have assumed that the benchmarks have beenimplemented as multi-threaded shared-memory programs withdata imperfectly placed, requiring communication through the on-chip interconnect. We compare our nanophotonic architecture to amany-core electrically-interconnected alternative system, for whichwe assume an on-chip mesh network, power-limited to 50 W, andan electrical connection to memory with bandwidth limited by thepin count and pin bandwidth anticipated by ITRS in 2017. Ourestimated results for the 17 nm technologynode are shown in Table 2. The final columnlists the modeled ratio of optical performanceto electrical performance per unit of dissipat-ed heat. Note that four of the benchmarksshow a factor-of-20 improvement fornanophotonics over wires. The other twobenchmarks do not show significantimprovements because they are not band-width constrained.

7. ConclusionThe many-core architecture presented here— with the cores divided into silicon com-pute clusters, connected to each other and tooff-chip memory using nanophotonic tech-nology — will continue to evolve [3] as wefurther explore the implications of a highlyparallel interconnect for the programmer.We believe that the use of DWDM in inte-grated-circuit interconnects is inevitable,and that the optical components that wedescribe here are essential elements of thatapproach. The potentially high bandwidth ofan optical interconnect in general-purposemany-core processors will be significantlycompromised if an electronically-reconfig-urable circuit-switched mesh or torus archi-tecture [5] (essentially a photonic implemen-tation of today’s copper-wire global intercon-nects) is employed. Instead, we propose to

build an all-optical arbitration system that relies on the samenanophotonic building blocks as the data-transmission network,allowing the transmitters themselves to determine whether areceiver is available, and to begin sending in only a few clock cycles.We have modeled the performance of this system using the HPCCbenchmarks, and we have found that a performance increase of 20over a purely electronic interconnect — essentially 4–5 “extra”Moore’s Law generations — is possible for applications that heavi-ly use the interconnect. This extraordinary performance boost is acritical goal for those of us advocating such a radical departure fromcurrent semiconductor engineering practice: the transition to thisnew interconnect technology will be so painful for the IT industrythat only an order-of-magnitude improvement in compute band-width will make the risk and effort worthwhile.

References[1] http://www.itrs.net/.[2] R. G. Beausoleil, P. J. Kuekes, G. S. Snider, S.-Y. Wang, ,

and R. S. Williams, “Nanoelectronic and NanophotonicInterconnect (Invited Paper),” Proc. IEEE 96, 230–247(2008).

[3] D. Vantrease, R. Schreiber, M. Monchiero, M. McLaren, N.P. Jouppi, M. Fiorentino, A. Davis, N. Binkert, R. G.Beausoleil, and J. Ahn, “Corona: System Implications ofEmerging Nanophotonic Technology,” in Proceedings of the35 International Symposium on Computer Architecture(ISCA 2008) (Beijing, China, 2008). To appear.

22leos03.qxd 6/16/08 1:16 PM Page 21


[4] N. Kirman, M. Kirman, R. K. Dokania, J. Martínez, A. B.Apsel, M. A. Watkins, and D. H. Albonesi, “OpticalTechnology in Future Bus-based Multicore Designs:Opportunities and Challenges,” IEEE Micro 27, 56–66(2007).

[5] K. Bergman and L. Carloni, “Power efficient photonic net-works on-chip,” in Proc. Soc. Photo-Opt. Instrum. Eng.,vol. 6898, p. 689813 (2008).

[6] R. Ho, On Chip Wires: Scaling and Efficiency, Ph.D. thesis(2003).

[7] W. M. Green, M. J. Rooks, L. Sekaric, and Y. A. Vlasov,“Ultra-compact, low RF power, 10 Gb/s silicon Mach-Zehnder modulator,” Opt. Express 15, 17106–17113(2007).

[8] Y.-H. Kuo, Y.-K. Lee, Y. Ge, S. Ren, J. E. Roth, T. I.Kamins, D. A. B. Miller, and J. S. Harris, “Strong quantum-confined Stark effect in germanium quantum-well struc-tures on silicon,” Nature 437, 1334–1336 (2005).

[9] J. E. Roth, O. Fidaner, R. K. Schaevitz, Y.-H. Kuo, T. I.Kamins, J. S. Harris, and D. A. B. Miller, “Optical modula-tor on silicon employing germanium quantum wells,” Opt.Express 15, 5851–5859 (2007).

[10] A. Liu, R. Jones, L. Liao, D. Samara-Rubio, D. Rubin, O.Cohen, R. Nicolaescu, and M. J. Paniccia, “A high-speed sil-icon optical modulator based on a metal-oxide-semiconduc-tor capacitor,” Nature 427, 615–618 (2004).

[11] A. Bhatnagar, C. Debaes, H. Thienpont, and D. A. B.Miller, “Receiverless detection schemes for optical clock dis-tribution,” in Proc. Soc. Photo-Opt. Instrum. Eng., vol.5359, pp. 352–359 (2004).

[12] Q. Xu, B. Schmidt, S. Pradhan, and M. Lipson,“Micrometre-scale silicon electro-optic modulator,” Nature435, 325–327 (2005).

[13] A. Kovsh, I. Krestnikov, D. Livshits, S. Mikhrin, J. Weimert,

and A. Zhukov, “Quantum dot laser with 75nm broad spec-trum of emission,” Opt. Lett. 32, 793–795 (2007).

[14] A. Gubenko, I. Krestnikov, D. Livshtis, S. Mikhrin, A.Kovsh, L. West, C. Bornholdt, N. Grote, and A. Zhukov,“Error-free 10 Gbit/s transmission using individual Fabry-Perot modes of low-noise quantum-dot laser,” Electron. Lett.43, 1430–1431 (2007).

[15] http://www.innolume.com/. [16] B. R. Koch, A. W. Fang, O. Cohen, and J. E. Bowers,

“Mode-locked silicon evanescent lasers,” Opt. Express 15,11225–11233 (2007).

[17] S. Xiao, M. H. Khan, H. Shen, and M. Qi, “A high-ly compact third-order silicon microring add-dropfilter with a very large free spectral range, a flat pass-band and a low delay dispersion,” Opt. Express 15,14765–14771 (2007).

[18] M. S. Nawrocka, T. Liu, X. Wang, and R. R. Panepucci,“Tunable silicon microring resonator with wide free spectralrange,” Appl. Phys. Lett. 89, 071110 (2006).

[19] Q. Xu, S. Manipatruni, B. Schmidt, J. Shakya, and M.Lipson, “12.5 Gbit/s carrier-injection-based silicon micro-ring silicon modulators,” Opt. Express 15, 430–436 (2006).

[20] Q. Xu, B. Schmidt, J. Shakya, and M. Lipson,“Cascaded silicon micro-ring modulators for WDMoptical interconnection,” Opt. Express 14,9430–9435 (2006).

[21] A. W. Fang, R. Jones, H. Park, O. Cohen, O. Raday, M. J.Paniccia, and J. E. Bowers, “Integrated AlGaInAs-siliconevanescent race track laser and photodetector,” Opt. Express15, 2315–2322 (2007).

[22] Q. Xu, D. Fattal, and R. G. Beausoleil, “Silicon microringresonators with 1.5-µm radius,” Opt. Express 16,4309–4315 (2008).

[23] http://icl.cs.utk.edu/hpcc/hpcc_results.cgi.

VViissiitt tthhee LLEEOOSS wweebb ssiittee ffoorr mmoorree iinnffoorrmmaattiioonn::

wwwwww..ii--LLEEOOSS..oorrgg

22leos03.qxd 6/16/08 1:16 PM Page 22


22leos03.qxd 6/16/08 1:16 PM Page 23


News

Call for Nominations:The 2009 Tyndall AwardNominations are now being accepted for the 2009 TyndallAward, which will be presented at OFC/NFOEC 2009.

This award, which is jointly sponsored by the IEEELasers and Electro-Optics Society and the Optical Society ofAmerica, is presented to a single individual who has madeoutstanding contributions in any area of lightwave technol-ogy, including optical fibers and cables, the optical compo-

nents employed in fiber systems, as well as the transmissionsystems employing fibers. With the expansion of this tech-nology, many individuals have become worthy of considera-tion. The deadline for nominations is 10 August, 2008.

For more information contact [email protected] orcheck the LEOS web for more details: www.i-leos.org –click “Awards” tab.

Nominations for IEEE Medals and RecognitionsThe IEEE Awards Board is seeking nominations for IEEEMedals and Recognitions, and encourages the use of its onlinePotential Nominee Form. This form allows a preliminary reviewof a nominee by the selection committee and an opportunity toobtain feedback prior to submitting an official nomination form.The Potential Nominee Form is available on the IEEE

Awards Web Page at http://www.ieee.org/portal/pages/about/awards/noms/potnomform.html.The deadline for submission of an official nomination formfor any of the IEEE Medals and Recognitions is 1 July,2008. For questions concerning the Potential NomineeForm, please contact [email protected].

Call for Nominations:2009 IEEE/LEOS Young Investigator AwardThe IEEE LEOS Young Investigator Award was establishedto honor an individual who has made outstanding techni-cal contributions to photonics (broadly defined) prior to hisor her 35th birthday.

The award shall consist of a certificate of recognitionand an honorarium of $1,000. The funding for this awardis being sponsored by General Photonics Corporation.

Nomination packages will be due at the LEOS executiveoffice by 30 September, 2008. Nominees must be under 35years of age on Sept. 30th of the year in which the nomina-tion is made. The award may be presented either at theOptical Fiber Communications Conference (OFC), or the

Conference on Lasers and Electro-Optics (CLEO), to beselected by the recipient. The first award was CLEO 2007.

Nomination packages consist of a nomination cover page, astatement of the nominee’s research achievements in photonics,the nominee’s curriculum vitae, and three to five reference let-ters (to be received at the LEOS office prior to the deadline).

Please consider nominating an under-age-35 colleaguefor the inaugural cycle of this award!

For full information about the LEOS awards programlook under the “Awards” tab on the LEOS web site(http://www.i-leos.org/)

Nomination form can be found on page 26.

22leos03.qxd 6/16/08 1:16 PM Page 24


News (cont’d)

22leos03.qxd 6/16/08 1:16 PM Page 25


News (cont’d)

22leos03.qxd 6/16/08 1:16 PM Page 26

News (cont’d)

Dr. Kristina M. Johnson, provost andsenior vice president for AcademicAffairs at Johns Hopkins University,recently received the John Fritz Medalfrom the American Association ofEngineering Societies (AAES). She is thefirst woman so honored. Johnson wasone of seven honorees during the AAES29th annual awards ceremony in theGreat Hall of the National Academy ofEngineering on 5 May. She was cited forher internationally acknowledgedexpertise in optics, optoelectronic

switching and display technology.The John Fritz Medal, referred to as

the highest award in the engineeringprofession, is presented each year for sci-entific or industrial achievement in anyfield of pure or applied science. It wasestablished in 1902 as a memorial to thegreat engineer whose name it bears. Pastrecipients include Alexander GrahamBell (1907), Thomas Edison (1908),Alfred Nobel (1910), Orville Wright(1920) and Guglielmo Marconi (1923).

Johnson is an IEEE Fellow, LEOS

member, and Electrical Engineer who, asthe former dean of engineering at DukeUniversity, increased the engineeringfaculty by 50 percent, tripled the size ofthe teaching and research facilities, andtripled the number of women engineer-ing faculty, many in leadership positions.She co-founded the Colorado AdvancedTechnology Institute for Excellence inOptoelectronics and started several com-panies that are commercially successfulin color projection devices and intellec-tual property licensing.

IEEE Fellow First Woman Honored with John Fritz Medal

The changes introduced this year in thePhAST program directly address the maingoal through all of the objectives, but par-ticularly through objectives 3 and 4.

PhAST is strongly focused on applica-tions that are market driven (so addressingobjective 3), with a program intended topresent the latest technologies and thecontext in which they are deployed inapplications.

Addressing objective 4 is more challeng-ing. In practice, increased co-operation withother disciplines implies developing rela-tionships with other societies, a process thattakes time. LEOS is at the center of a num-ber of ‘joined up’ networks of societies.These networks include communications,displays, nanotechnology, biotechnologyand biometrics, some of which have operat-ed successfully for many years, with others,emerging only recently emerged.CLEO/QELS/PhAST falls in the former cat-egory and is owned and run by three co-sponsoring societies: LEOS, OSA and APS.The OIDA is involved in developing aspectsof the PhAST program for the first time.

The OIDA involvement is bringing acompletely new perspective to PhASTthrough the Business Development sessions.These sessions are aimed at developing anunderstanding of the entire product valuechain, and developing roadmaps for the rel-evant industries going forward. This is

being achieved through a mixture of presen-tations by senior managers and technolo-gists followed by interactive panel sessionsto develop further the ideas that emergeduring the presentations.

The OIDA part of the program is fur-thermore focused on the emerging area of‘green’ photonics. As well as resonatingwith the interests of LEOS, this also forms alink with IEEE as a whole, which hasdeclared sustainable technologies a priorityarea. It is becoming clear that the photonicsindustry will provide many solutions forreducing environmental pollution and theworld’s carbon footprint. These solutionsinclude highly efficient solid-state lighting,effluent-free laser processing, new low ener-gy display technologies and photovoltaicenergy generation.

Other key parts of the program includethe Power Lunch, organized by MiltonChang, and the PhAST/Laser Focus WorldInnovation Awards.

PhAST raises a number of questions forLEOS. The principal question is how readi-ly the format could be used elsewhere. LEOSis already working with OIDA to exploreextending the concept of matching technicaland business sessions in a single workshop.LEOS is also actively working with otherIEEE societies to identify more applications-focused topics that use photonics. A secondpoint is that this year’s PhAST sessions are

entirely invited, which is appropriate forcertain meetings but clearly imposes limitson member participation. A third point isthat there is no charge to attend PhAST.Many attendees will have registered forCLEO/QELS, but an intention of PhAST isto support the CLEO Exhibition by attract-ing walk-ins. Clearly having no registrationfee is not an option for independent events.My belief is that we can develop a model toaddress applications focused topics thatcombines the LEOS core values of volunteerservice, education, integrity, technical rigorand inclusiveness, and at the same time issustainable.

All of these questions will be addressedby LEOS over the coming months. The ulti-mate aim of LEOS is to serve its membersand the photonics community. Moving intomore areas of application is one way ofaddressing this aim and I would welcomeyour views on how this can best be achieved.

Finally, I would like to remind you of theappeal I made in January where I invitedyou to recruit one new member each in2008. As the middle of the year approaches,LEOS membership has started to grow forthe first time in several years. This is excel-lent news, but there is still much that couldbe done. The best recommendation is a per-sonal one – if you value LEOS membership,please spread the message to your colleaguesand build a stronger LEOS community.

President’s Column(continued from page 3)

22leos03.qxd 6/16/08 1:16 PM Page 27


News (cont’d)

The fiber-optics community haslost a great icon with the passingof Prof. Pak Chu after a one yearbattle with cancer. His passing wascommemorated with an extendedservice of thanksgiving, eulogy,tributes, and valediction at theWest Sydney Chinese ChristianChurch that was attended by anoverflowing congregation of col-leagues, friends, and fellowChristians.

Prof. Chu is a great championand a pioneer in the field of fiberoptics in Australia. He set up thevery first academic fiber-opticsresearch laboratory in Australia atthe University of New South Wales(UNSW) in 1970s. The laboratorywas equipped with the first pre-form-making and fiber-drawingfacilities in the country. For thelast 30 years, Prof. Chu has madeextensive contributions to theoptical fiber technology. In early1980s, he invented an acclaimedmethod to measure the refractive-index profiles of fiber preforms.The method was adopted in anadvanced commercial instrumentfor the characterization of opticalfibers. In the mid-1980s, he suc-cessfully solved the problem ofsoliton interaction, which laid animportant foundation for solitoncommunications, a field in whichhe maintained an interest for over20 years. In addition, Prof. Chuhad worked on a large number ofinnovative optical devices, includ-ing nonlinear optical switches,fiber-optic acoustic sensors, poly-mer fiber Bragg gratings, and ver-

tical optical polymer waveguidecouplers. This work has attractedconsiderable attention over theyears. His study on chaos commu-nications is also very influential. Inrecent years, he spent much effortin promoting polymer optical fibertechnology, a growing field withmany potential applications inshort-distance transmission andoptical sensing. His contributionsto the field of fiber optics areinvaluable.

On the teaching side, Prof. Chusupervised over 40 doctoral andnumerous honours students anddeveloped the first undergraduatecourses in optical communicationstechnology in Australia.

Prof. Chu was a popular speaker,particularly at meetings inSoutheast Asia, giving numerouscontributed and invited talks aswell as short courses and tutorials.He was a Fellow of several scientificsocieties including the Australian

Technological Society and theOptical Society of America. Hepublished almost 500 journal andconference papers, and was hon-oured in 2001 with the CentenaryMedal of Australia for his contribu-tion to optical communications.

In 2001, Prof. Chu retired fromUNSW and moved to the CityUniversity of Hong Kong (CityU)where he became Professor andDirector of the OptoelectronicsResearch Centre (RCO). During hisDirectorship of RCO, he made sig-nificant contributions to the devel-opment of the photonics researchprograms at CityU. His cheerful,positive, and supportive tempera-ment enabled him to be an excel-lent mentor and a real friend ofmany people. Prof. Chu was alsoactive in fostering collaborationswith industry and organising inter-national conferences and trainingworkshops to raise the profile of thephotonics community in HongKong. After his retirement fromCityU in 2006, he joined a localcompany, LINKZ InternationalLimited (the former NetworkingCable Business Unit of LTK),where he continued to make contri-butions in photonics for the rest ofhis time in Hong Kong.

Prof. Chu is survived by his wifeEva, daughter Evelyn, son Desmondand five grandchildren.

Optoelectronics Group,Department of Electronic

Engineering,City University of Hong Kong,

April 2008

Memoriam: Professor Pak Lim Chu12 November 1940 – 15 March 2008

Prof. Chu at the City University ofHong Kong

22leos03.qxd 6/16/08 1:16 PM Page 28


Career Section

The Quantum Electronics Award isgiven to honor an individual (orgroup of individuals) for outstandingtechnical contributions to quantumelectronics, either in fundamentals orapplication or both. The Award maybe for a single contribution or for adistinguished series of contributionsover a long period of time. No candi-date shall have previously received amajor IEEE award for the same work.Candidates need not be members ofthe IEEE or LEOS. The deadline fornominations is 16 February.

Jeffrey H. Shapirois Director of theResearch Laboratoryof Electronics (RLE)at the MassachusettsInstitute of Tech-nology (MIT). Hereceived the S.B.,S.M., E.E., andPh.D. degrees in

Electrical Engineering from MIT in1967, 1968, 1969, and 1970, respec-tively. As a graduate student he was aNational Science Foundation Fellow, aTeaching Assistant, and a Fannie andJohn Hertz Foundation Fellow. Hisdoctoral research was a theoreticalstudy of adaptive techniques forimproved optical communicationthrough atmospheric turbulence.

From 1970 to 1973, Dr. Shapirowas an Assistant Professor of

Electrical Sciences and AppliedPhysics at Case Western ReserveUniversity. From 1973 to 1985, hewas an Associate Professor ofElectrical Engineering at MIT, and in1985, he was promoted to Professorof Electrical Engineering.

From 1989 until 1999 Dr. Shapiroserved as Associate Department Headof MIT's Department of ElectricalEngineering and Computer Science.In 1999 he became the Julius A.Stratton Professor of ElectricalEngineering. In 2001, Dr. Shapirowas appointed Director of RLE.

Dr. Shapiro's research interestshave centered on the application ofcommunication theory to optical sys-tems. He is best known for his workon the generation, detection, andapplication of squeezed-state lightbeams, but he has also publishedextensively in the areas of atmospher-ic optical communication, coherentlaser radar, and quantum informationscience.

Dr. Shapiro is a fellow of theInstitute of Electrical and ElectronicsEngineers, of the Optical Society ofAmerica, of the American PhysicalSociety, and of the Institute of Physics,and he is a member of SPIE (TheInternational Society for OpticalEngineering). He has been an AssociateEditor of the IEEE Transactions onInformation Theory and the Journal ofthe Optical Society of America, and was

the Principal Organizer of the SixthInternational Conference on QuantumCommunication, Measurement andComputing (QCMC'02). He currentlyco-chairs the Steering Committee forthe International Conferences onQuantum Communication, Measure-ment and Computing, and is Co-Director of the W. M. KeckFoundation Center for ExtremeQuantum Information Theory.

Horace P. Yuen is aProfessor of ElectricalEngineering andComputer Scienceand Professor ofPhysics and Astro-nomy at North-western University.He received his

degrees in Electrical Engineering fromMassachusetts Institute of Technology.His technical research interests aremainly in the areas of communicationand cryptography, especially thosewith quantum effects. He is a recipientof the first International QuantumCommunication Award presented byTamagawa University of Japan, a fel-low of the American Physical Society,and a senior member of IEEE. Several ofhis papers are collected in various specialvolumes, including "One HundredYears of Physical Review," which waspublished by the American PhysicalSociety in 1993.

2008 IEEE/LEOS Quantum Electronics AwardRecipients: Jeffrey H. Shapiro and Horace P. Yuen

Jeffrey H.Shapiro

Horace P. Yuen

22leos03.qxd 6/16/08 1:16 PM Page 29


Career Section (cont’d)

The Distinguished Lecturer Awards are presented to honor interesting speakers who have made recent significant contributionsto the field of lasers and electro-optics. The Distinguished Lecturers speak at LEOS Chapters worldwide. Please contact you localchapter to see when one of the Lecturers will be speaking in your area. A list of current LEOS Chapter Chairs is available on theLEOS web site (www.i-LEOS.org).

This year’s new Lecturers are:

John C. Cartledge, Queen’s University, Ontario, CanadaTopic of Lecture: Optical and Electronic Signal Processing for Fiber-Optic Communications

John M. Dudley, Laboratoire d’Optique P.M. Duffieux, FranceTopic of Lecture: Nonlinear Fiber Optics and Supercontinuum Generation: New Fibers, New Opportunities

Prem Kumar, Northwestern University, IL, USATopic of Lecture: Fiber-Optic Quantum Communications and Information Processing

Lecturers serving a second term are:

Weng Chow, Sandia National Laboratory, NM, USATopic of Lecture: Many-Body Effects and Their Influences on Semiconductor Lasers

Silvano Donati, Universita Pavia, ItalyTopic of Lecture: Coupling in Lasers and Applications to Self-Mixing Interferometry and Chaotic Cryptography

El-Hang Lee, INHA University, South KoreaTopic of Lecture: Optical Printed Circuit Board (O-PCB) and VLSI Photonics

Colin McKinstrie, Bell Labs, Alcatel-Lucent, NJ, USATopic of Lecture: Optical Signal Processing by Parametric Devices

Cun-Zheng Ning, Arizona State University, AZ, USATopic of Lecture: Nanolasers Based on Semiconductor Nanowires and Surface Plasmons

2008-09 Distinguished Lecturers:

John C. Cartledge was a Member ofthe Scientific Staff at Bell-NorthernResearch, Ottawa, Ontario, Canada,from 1979 to 1982, where his workinvolved fiber-optic systems for theexchange access network and high-capacity digital radio systems. Since1982 he has been with theDepartment of Electrical andComputer Engineering, Queen's

University. In July 2002 he was appointed an inauguralrecipient of a Queen’s Research Chair. He has spent one-year

sabbatical leaves with the Lightwave Systems TechnologyResearch Division of Bellcore, Red Bank, NJ, in 1988-89,and with the Optical Communications Department of TeleDanmark Research, Hørsholm, Denmark in 1995-96. Dr.Cartledge has served as a consultant and instructor in thearea of lightwave technology to several organizations. Hiscurrent research interests include optical modulators, opti-cal signal processing (wavelength converters, optical regen-erators), electronic signal processing for arbitrary opticalwaveform generation, and dispersion compensating fiberBragg gratings. Dr. Cartledge is Chair of the OpticalNetworks and Systems Committee, IEEE Lasers and

2008-2009 LEOS Distinguished Lecturers

22leos03.qxd 6/16/08 1:16 PM Page 30



Electro-Optics Society, 2007-2009, and is a Fellow of theOptical Society of America.

Dr. Cartledge’s conference organization activitiesinclude serving on the Technical Program Committees forthe Conference on Optical Fiber Communications (1994-1997 and 2005-2006), the IEEE LEOS Annual Meeting(2002-2008), the European Conference on OpticalCommunications (2006-2008), and the joint meeting ofthe International Conference on Integrated Optics andOptical Fibre Communications and the EuropeanConference on Optical Communications (1997). He was aProgram Chair for the joint meeting of the Conference onOptical Fiber Communications and the National FiberOptics Engineers Conference (2008).

Title of Talk: Optical and Electronic Signal Processing forFiber-Optic Communications

Currently, there is a substantial research and develop-ment effort directed toward optical signal processing andelectronic signal processing for fiber-optic communica-tions. Much of the work is aimed at mitigating theeffects of transmission impairments such as chromaticdispersion, polarization mode dispersion, fiber nonlin-earities, amplifier noise, and band-limiting. It is beingpursued in the presence of an ongoing interest in increas-ing the per-channel bit rate in order to meet the grow-ing demand for telecommunication services. For opticalsignal processing, a variety of approaches are availablefor implementing functions such as 3R (re-amplifica-tion, reshaping and re-timing) regeneration. For elec-tronic signal processing, advanced digital signal process-ing is being applied to both fiber-optic transmitters andreceivers, and has lead to a renewed interest in coherentoptical detection. This lecture presents an overview ofoptical and electronic signal processing technologies,including a critical assessment of each approach. Specificexamples are considered in more detail to highlight keyaspects of the technologies.

John Dudley received B.Sc andPh.D. degrees from the Universityof Auckland in 1987 and 1992respectively. In 1992 and 1993, hecarried out postdoctoral researchat the University of St Andrews inScotland before taking a lecturingposition in 1994 at the Universityof Auckland. In 2000, he wasappointed Professor at the

University of Franche-Comté in Besançon, France,where he currently heads the Optoelectronics andPhotonics research group. He was named a member of

the Institut Universitaire de France in 2005, and elect-ed a Fellow of the Optical Society of America and aSenior Member of the IEEE in 2007. He serves on theEditorial boards of Optics Express, Optical andQuantum Electronics and Optical Fiber Technology. Heis General Chair of CLEO Europe 2009 and currentlyserves as the secretary of the Quantum Electronics andOptics Division of the European Physical Society.

Title of Talk: Nonlinear Fiber Optics and SupercontinuumGeneration: New Fibers, New Opportunities

Research in nonlinear fiber optics is currently under-going dramatic expansion, motivated by advances anddevelopments in new classes of optical fiber and theready availability of a wide range of optical pumpsources. This lecture will survey selected recent work inthis field that has investigated novel nonlinear propaga-tion effects in both photonic crystal and highly nonlin-ear optical fibers, and will focus particular attention onthe physics and applications of supercontinuum genera-tion. The lecture will provide both a tutorial review ofthe basic supercontinuum generation broadening mech-anisms as well as a discussion of recent developmentsand applications.

Prem Kumar is the AT&TProfessor of Information Technologyin the department of ElectricalEngineering and Computer Scienceand Director of the Center forPhotonic Communication andComputing in the McCormickSchool of Engineering and AppliedScience at Northwestern University.He also holds an appointment as

Professor of Physics and Astronomy and joinedNorthwestern in 1986 after spending five years at MIT asa research scientist. He received a Ph.D. in Physics fromthe State University of New York at Buffalo in 1980. Heis the author or co-author of over 400 publications,including one edited book, six patents, over 140 papers inpeer-reviewed journals, 40 articles in hard-bound vol-umes, and over 80 invited conference papers. His researchfocuses on the development of novel free-space and fiber-optic devices for ultrahigh-speed optical and quantumcommunication networks. He is a fellow of the OpticalSociety of America (OSA), a fellow of the AmericanPhysical Society (APS), a fellow of the Institute ofElectrical and Electronic Engineers (IEEE), and a fellow ofthe Institute of Physics, UK (IoP). In 2006 he received theMartin E. and Gertrude G. Walder Research ExcellenceAward from Northwestern University and in 2004 he was

22leos03.qxd 6/16/08 1:16 PM Page 31



the recipient of the 5th International QuantumCommunication Award established by the TamagawaUniversity in Tokyo, Japan. On the academic side, his pro-fessional services have included Associate Editor, OpticsLetters; General (Program) Co-Chair, QELS 2008(2006); Principal Organizer, 4th InternationalConference on Quantum Communication, Measurement,and Computing, 1998. On the business side, he is thefounder of NuCrypt LLC, a startup company focusing onthe commercialization of quantum encryption technologyfor securing the physical layer of fiber-optic and free-spaceoptical networks.

Title of Talk: Fiber-Optic Quantum Communications andInformation Processing

Recognizing the ubiquitous standard optical fiberfor long-distance transmission and the widespreadavailability of efficient active and passive fiber devices,there are significant efforts underway to develop practi-cal resources for quantum communications and infor-mation processing in optical fiber networks.Entanglement, which refers to the nonclassical depend-ency of physically separated quantum systems, is onesuch resource that is essential for implementing manyof the novel functions of quantum information process-ing. Therefore, the efficient generation and distribu-tion of quantum entanglement in fiber optical systemsis of prime importance. Entanglement has been histor-ically produced by use of the spontaneous parametricdown-conversion process in second-order nonlinearcrystals, wherein one higher-frequency pump photonsplits into two lower-frequency daughter photonswhich can be entangled. Coupling such down-convert-ed photons into optical fibers without degradingentanglement, however, has remained a challengingtask. Fortunately, the prospects for ready availability ofentanglement in the telecom band have dramaticallyimproved in the last few years by the emergence of anew technique for generating entanglement directly inthe fiber itself. This technique utilizes the Kerr non-linearity of standard optical fiber to produce quantumcorrelated photons through the spontaneous four-wavemixing process. The correlated photons can be entan-gled in various ways by incorporating indistinguishablepathways in the four-wave mixing amplitude. In thislecture, I will review the status of this field by describ-ing recent experiments that demonstrate the generationand distribution of quantum entanglement is wave-division multiplexed optical fiber systems. I will alsopresent some recent results on utilizing such entangle-ment for quantum communications and informationprocessing tasks.

Weng Chow received the Ph.D.degree in physics from the Universityof Arizona. His dissertation workinvolved fluctuation phenomena inquantum optics. At present, he isDistinguished Member of theTechnical Staff at Sandia NationalLaboratories. Weng Chows primaryresearch interest is in the applicationof microscopic theory to semiconduc-

tor laser device development. Some of this work is described intwo texts, Semiconductor-Laser Physics and Semiconductor-Laser Fundamentals: Physics of the Gain Materials. His otherinterests include laser gyros, phased arrays, coupled lasers, quan-tum optics and optical ignition of pyrotechnics.

Weng Chow also holds the positions of Research Professor ofPhysics at Texas A&M University, Adjoint Professor of OpticalSciences at the University of Arizona and Honorary Professor ofPhysics at Cardiff University. He has served on the CLEO semi-conductor laser program committee and is an associate editor ofIEEE Journal of Quantum Electronics. Weng Chow is a fellowof the Optical Society of America, and recipient of the Dept ofEnergy, Basic Energy Science/Material Science Award, theAlexander von Humboldt Senior Scientist Award and the LEOSDistinguished Lecturer Award.

Talk title: Many-Body Effects and Their Influences onSemiconductor Lasers

The incorporation of many-body effects into semiconductor-laser gain calculations has led to significant improvement inaccuracy and predictive capability. This talk will explain themany-body effects and sketch the formulation of a laser theorycontaining the necessary physics. Examples of application of thetheory will be presented. These examples involve laser systemsranging from VCSELs to wide-bandgap and quantum-dot lasers.

S. Donati is full Professor atUniversity of Pavia since 1980. Heauthored 2 books (Photodetectors,Prentice Hall 1999, and Electro-optical Instrumentation, PrenticeHall 2004), about 250 papers inJournal and Conference Proceedings,and has been the Guest Editor of adozen Special Issues (JSTQE, JoO-A,Opt.Engineer., JQE, etc.). His sem-

inal papers on self-mixing interferometry and opticalchaotic cryptography have totaled 500+ citations. He is aFellow of IEEE and of OSA. Has founded and has beenthe Chair of the LEOS Italian Chapter. Has been LEOSVP Region 8 membership and BoG member of the LEOS.He is presently the Chair of the IEEE Italy Section.

22leos03.qxd 6/16/08 1:17 PM Page 32



Title of talk: Coupling in Lasers and Applications toSelf-Mixing Interferometry and Chaotic Cryptography

In this presentation, we start with a brief theoretical intro-duction to mutual and self coupling phenomena in laser oscil-lator, and then describe in details two applications. The firstis self-mixing interferometry for measurements of displace-ment, distance, vibration, and angle, and physical parameterslike coupling factors, line width, alfa-factor. In this case, thelaser undergoes self-injection at weak level, leading to anamplitude and frequency modulation driven by external opti-cal path. The second application is optical chaos, which is gen-erated by the laser source at strong level of injection. Wedescribe mutual and self-injection generation of chaos, and thefirst step of development to cryptography, that is, synchro-nization. Then, we will review several schemes of coding anddecoding of information, i.e., chaotic masking and CSK (chaosshift keying) and how they can be implemented, along withtheoretical and experimental results carried out recently.

El-Hang Lee obtained his B.S.E.E.(summa cum laude), at SeoulNational University, Korea, 1970;M.S., M.Phil., and Ph.D., AppliedPhysics, at Yale University, 1973,1975 and 1977, respectively, underProf. John. B. Fenn (Yale NobelLaureate, Chemistry, 2002) andProf. Richard. K. Chang (HenryFord II Professor, former student of

Prof. N. Bloembergen, Harvard Nobel Laureate, Physics,1981). He conducted teaching, research and management atYale, Princeton, MEMC, AT&T Bell, ETRI (vice president),KAIST, and at INHA in the fields of semiconductorphysics, materials, devices, optoelectronics, photonics, andoptical communication. Founding Dean, School ofCommunication and Information Engineering; Dean,Graduate School of the Information Technology andTelecommunications; Founding Director, OPERA (Opticsand Photonics Elite Research Academy) and m-PARC(micro/nano-Photonics Advanced Research Center); VicePresident, Optical Society of Korea; Founding President,IEEE-LEOS Korea; Founding Director, SPIE-Korea. 230international refereed SCI-covered journal and reviewpapers; 640 international conference presentations; 100 ple-nary, keynote, and invited talks in international confer-ences; Edited books and international proceedings; 120international patents; 100 services as international confer-ence chair, committee member, and advisor. Fellow, IEEE(USA), IEE (UK), OSA (USA), SPIE (USA), APS (USA),KPS (Korea), IEEK (Korea), and Life Fellow, KoreanAcademy of Science and Technology. Recipient of 15national and international awards, including Presidential

Medal of Honor (Science), Korea; King Sejong Award(Science), Korea; IEEE/LEOS Chapter-of-the-Year Award,and the IEEE Third Millennium Medal, USA.

Title of Talk: Optical Printed Circuit Board (O-PCB) andVLSI Photonics

This lecture presents a comprehensive review andoverview on the cutting-edge frontier science andengineering of micro/nano-photonic integration thatwe have been pursuing aiming for what we call “opti-cal printed circuit boards” (O-PCBs) and VLSI pho-tonic chips. It discusses on the theory, design, fabrica-tion, and integration of micro/nano-scale photonicdevices, circuits, and networks in the form of “VLSIphotonic integrated circuits” (VLSI-PICs) and “opti-cal micro/nano-networks (O-MNNs)” of generic andapplication-specific nature on O-PCB platforms.These systems are designed to be compact, high-speed, light-weight, low-powered, low-cost, intelli-gent, and environmentally friendly as applicable fordatacom, telecom, transportation, aero-space, avion-ics, bio/medical, sensor, and environmental systems.The O-PCBs, VLSI-PICs and O-MNNs process opticalsignals through optical wires, devices, and circuits incontrast to the traditional electrical PCBs, VLSI-ICs,and networks. The O-PCBs and VLSI photonic sys-tems are to overcome the limitations of the electricalPCBs and VLSI-IC systems and are also aimed to inte-grate convergent IT/BT/NT micro/nano-devices, cir-cuits, and chips for broad based applications andusages. The new optical systems consist of 2-dimen-sional planar arrays of optical wires, circuits anddevices of micro/nano-scale to perform the functionsof sensing, storing, transporting, processing, switch-ing, routing and distributing optical signals on flatmodular boards or chips. The integrated optical com-ponents include micro/nano-scale light sources, wave-guides, detectors, switches, modulators, sensors,directional couplers, multi-mode interference devices,AWGs, wavelength filters, micro-ring resonatordevices, photonic crystal devices, plasmonic devices,and quantum devices, made of polymer, silicon andother semiconductor materials. The lecture discussesscientific and technological issues, challenges, andprogresses regarding the miniaturization, interconnec-tion and integration of micro/nano-scale photonicdevices, circuits, and networks leading to ultra-smalland very large scale integration and discusses theirpotential applications. The issues include the diversecompatibility issues between micro/nano-devices suchas materials mismatch, size mismatch, shape mis-match, mode mismatch, optical mismatch, mechani-

22leos03.qxd 6/16/08 1:17 PM Page 33



cal/thermal mismatch and the micro/nano-opticaleffects such as micro-cavity effects, non-linear effects,and quantum optical effects in small devices. Scalingrules for the miniaturization and integration of themicro/nano-photonic systems will also be discussed incomparison with those of the electrical systems. Newphysics, new materials, new designs, and new visions,issues and challenges of the optical micro/nano-opticalcircuits, networks and systems will be discussed alongwith the historical perspectives of the electrical tech-nology. Recent progresses and examples will be pre-sented along with the future outlook.

Colin J. McKinstrie receivedBSc and PhD degrees from theUniversities of Glasgow andRochester, in 1981 and 1986,respectively. From 1985 to 1988he was a Postdoctoral Fellow ofLos Alamos National Laboratory,where he was associated with theApplied Physics Division andthe Center for Nonlinear

Studies. In 1988 he returned to the University ofRochester as a Professor of Mechanical Engineering anda Scientist in the Laboratory for Laser Energetics, wherehis main research interests were plasma-based particleacceleration, laser-plasma interactions and nonlinearfiber optics. Since 2001 Dr. McKinstrie has been aMember of the Technical Staff at Bell Laboratories,where his research concerns the amplification andtransmission of optical pulses in communication sys-tems. He has served on technical committees for CLEO,FiO, LEOS and OFC, and is the Chair of the OSAQuantum Electronics Division.

Title of Talk: Optical Signal Processing by ParametricDevices

Parametric devices based on four-wave mixing infibers provide many functions that are required byoptical communication systems. When operated in thelinear regime, parametric devices provide amplifica-tion, frequency conversion and phase conjugation, allwith high gain levels and broad bandwidths. They canalso be used to buffer, monitor and switch optical sig-nals. When operated in the nonlinear regime, para-metric devices regenerate signals. They also produceentangled and squeezed states of light. In this talkrecent research on parametric devices will bereviewed, and the implications of this research forclassical and quantal communication systems will bediscussed

Cun-Zheng Ning obtained hisPhD in Physics from Universityof Stuttgart. He was a SeniorScientist, group leader, or taskmanager at NASA Ames Centerfor Nanotechnology, NASAAmes Research Center from1997-2006. He joined ArizonaState University in 2006, wherehe is a Professor of Electrical

Engineering with the Center for Nanophotonics,Arizona Institute of NanoElectronics, and AffiliateProfessor in Materials and in Physics. Dr. Ning hasbeen conducting research in the general fields of laserphysics, semiconductor lasers, optoelectronic devicemodeling and simulation for the last 20 years.Recently, he has also developed a significant experi-mental program in semiconductor nanowire researchon nanowire growth, optical characterization, anddevice fabrication. His group was the first to growantimonide nanowires and first to demonstrate a sin-gle-nanowire infrared laser. His group has also beenactively engaged in semiconductor plasmonic devicesfocusing on integrating semiconductor nanostructureswith metallic nanostructures for various applicationssuch as nanolasers, solar cells and detectors. He haspublished 120 scientific papers and given many con-ference presentations including ~50 invited talks. Hehas served in many international conference commit-tees including SPIE Photonics West, OSA annualmeetings, and CLEO. He also served in editorial com-mittee or as a special topic editor for a few journalsincluding IEEE JQE, J. Special. Topics in QuantumElectron., J. Opt. Soc. Am., Optics Express, etc. Forhis research at NASA, he was winner of many NASAand NASA contractor awards.

Title of Talk: Nanolasers Based on Semiconductor Nanowiresand Surface Plasmons

The pursuit of nanotechnology in general and miniaturiza-tion of electronic devices in particular have seriously challengedthe optoelectronics community to develop ever smaller lasersand optoelectronic devices compatible with the trend in micro-electronics. Vertical-cavity surface emitting lasers measured afew microns were once the smallest lasers. The situation is nowrapidly changing over the last 5 years with the demonstration oflasing capability of a single semiconductor nanowire of ~ 100nanometers in diameter. The question of great importance is:what is the ultimate size limit of a laser?

To answer this and related questions, my lecture willstart with impressive recent progress in growth, fabrica-tion, and characterization of semiconductor nanowires and

22leos03.qxd 6/16/08 1:17 PM Page 34



demonstration of lasing activities in various wavelengths.These lasers represent the smallest lasers of any kind atpresent. We will show how this new type of miniaturizedlasers differs from the conventional semiconductor lasers.To further reduce the dimension of nanowire lasers, arecent proposal of using metal coating of semiconductorwires will be evaluated. We will show that a proper design

of a metal coated semiconductor nanowire can achieve las-ing threshold despite significant metal loss. Very recentexperimental results of demonstrating such lasers will bepresented. Finally some recent novel ideas involving sur-face plasmonic excitations at metal-semiconductor inter-face will be discussed where much smaller lasers could bepotentially made, with size independent of wavelengths.

Membership Section

There are many benefits to becoming an IEEE Senior Member:• The professional recognition of your peers for technical and professional excellence• An attractive fine wood and bronze engraved Senior Member plaque to proudly display.• Up to $25 gift certificate toward one new Society membership.• A letter of commendation to your employer on the achievement of Senior member grade

(upon the request of the newly elected Senior Member.)• Announcement of elevation in Section/Society and/or local newsletters, newspapers and notices.• Eligibility to hold executive IEEE volunteer positions.• Can serve as Reference for Senior Member applicants.• Invited to be on the panel to review Senior Member applications.

The requirements to qualify for Senior Member elevation are, a candidate shall be an engineer, scientist, educator, technical execu-tive or originator in IEEE-designated fields. The candidate shall have been in professional practice for at least ten years and shall haveshown significant performance over a period of at least five of those years.”To apply, the Senior Member application form is available in 3 formats: Online, downloadable, and electronic version. Formore information or to apply for Senior Membership, please see the IEEE Senior Member Program website:http://www.ieee.org/organizations/rab/md/smprogram.html

The following individuals were elevated to Senior Membership Grade thru March-May:

Timothy J. CarrigChing K. ChiaMarc P. ChristensenDennis DericksonGlen P. DudevoirChang-Hee LeeMark S. LeesonYiu-Wing LeungAnhui Liang

Alexander S. LogginovSergei A. MalyshevShahid MasudPhyllis R. NelsonJesper NorregaardDaniel L. PettersonAleksandar D. RakicJae-Hyun RyouMasatoshi Saruwatari

Hrayr A. SayadianMichael A. SeiglerJulian B. SooleBarton W. StuckAnna TzanakTimo AaltoGuido ChiarettiR. P. DahlgrenDavid L. Esquibel

Sarath D. GunapalaLawrence A. HornakJianjang HuangShafiqul IslamBoero L. PaoloYuan ShiJoseph J. TalghaderYuelin WangYong-Gang Zhang

Benefits of IEEE Senior Membership

New Senior Members

22leos03.qxd 6/16/08 1:17 PM Page 35


Membership Section (cont’d)

The 2007-08 academic year has beenanother busy year for the Central NewEngland Chapter of LEOS, colloquial-ly known as the Boston Chapter. Asthe oldest local LEOS Chapter in theIEEE, as well as one of its largest in acompact geographical area, with over300 local active members, we strive tomeet the high standards set by themany previous generations of pastlocal chapter chairs. To this end, wemaintain our chapter’s two strong tra-ditions of holding 10 monthly techni-cal seminars (Sept. – June) on currentresearch topics in Lasers/Optics withtruly outstanding speakers (see listbelow), and putting on a 10 speakertutorial Workshop on a single current“hot” topic, with 2 speakers per nighton 5 consecutive Wednesday nights(see list below) in the spring or fall.This year the workshop subject cho-sen was Plasmonics, which followedour previous years workshops onTerahertz Systems, Nanophotonics,and Photonic Crystals. We requirepre-registration for our yearly work-shop, and generally charge a minimalfee (e.g. $50 for all 5 nights) to coverthe cost of refreshments during thebreaks and printing of the handouts.Each year the number registering forour local workshop has increased, hit-ting 115 for this year’s fall PlasmonicsWorkshop. Equally rewarding, is thatwe are now receiving requests fromspeakers asking to give a talk at ourlocal yearly workshop once the topicis announced. This year we had toinclude a third speaker on one night,as it is just getting harder and harderto down-select from the abundance ofappropriate well-qualified speakersavailable.

Otherwise, on the mechanics ofrunning our chapter, we find thatmost of our yearly changes are more inthe category of fine-tuning, to betterrespond to our changing member’s

needs, although we did make perma-nent the two major changes we insti-tuted last year. First, we had movedour monthly meeting location to MITLincoln Laboratories in suburbanBoston, after alternatively meeting inthe old suburban Verizon / GTE Labsfor over 20 years (Verizon afterabsorbing GTE closed its labs andsold the land to a developer), and atBoston University in downtownBoston (BU began charging for park-ing at evening meetings). But, per-haps even this change might not befinal, as MIT Lincoln Laboratories ison Government property, we learnedthat we had no possible appeal, whenin December an approaching snow-storm led officials to close the build-ing we meet in, only hours before ourmeeting began. Second, we movedour meeting time to earlier in theevening and now serve free pizza andsoft drinks during the networkinghour before our monthly meetings.This was done because we learned,while being continuously re-educatedover the past 25 years that our indus-trial members will not drive intodowntown Boston traffic to pay forparking at a University location, thatfree pizza always attracts graduatestudents to our meetings at such alocation. Thus, we developed a hybridapproach to serve/attract at least partof the available student audience. Byserving pizza during the hourbetween when the MIT Lincoln Labworkday ends, and our monthly meet-ing starts, some of the large numberof MIT graduate students doing theirdaily research at Lincoln Lab delaytheir return to Cambridge for a coupleof hours, to eat pizza, talk with us,and attend our meeting, before catch-ing the last buses back to Cambridgeat 7:30 PM. Hopefully, those that stayin the Boston area after graduationwill continue to attend our meetings.

Already, a couple of recently graduat-ed, new-hire Lincoln Lab staff mem-bers, who volunteered to help run ourchapter, have greatly decreased theaverage age of our chapter officers.

In addition, we have also workeddiligently on maintaining, whatmight be thought of as a rather over-ambitious advertising campaign bythose from outside of the suburbanBoston area, of our monthly chaptermeetings and Workshops to promoteattendance. But, as we have learnedfrom much past bitter experience, weare competing with all the daily semi-nars and departmental colloquiums ofover 25 world-class Universities andColleges in the Boston/Cambridgearea, with their world-wide draw oftechnical speakers from every fieldin Physics, Engineering, andBiology. These day-time talks satu-rate the market for technical infor-mation for anyone associated with aUniversity, College, or academicresearch institution in the city.While the downside is that it isvery hard to convince an audienceof faculty members and researchersworking in the Boston/Cambridgearea that we can produce bettertalks in their respective field at ourevening chapter meeting, theupside is that we have the widestand deepest collection of talentfrom which to draw local speakers(we find that even out-of-countryspeakers, talking at a universityduring the day, are willing to giveanother talk that evening at thelocal LEOS chapter meeting).Hence, it is critical for us to designthe advertising for our chaptertalks to reach other audiences,which are mostly drawn from localindustry surrounding suburbanBoston, rather than from theresearch establishment at theBoston/Cambridge Universities.

LEOS Central New England Chapter (CNEC)

22leos03.qxd 6/16/08 1:17 PM Page 36


Membership Section (cont’d)

We reach our intended audience byutilizing the IEEE Boston Section/CNEC newsmagazine “The Reflector”,in both paper and electronic formats(paid circulation over 13,000 in easternMass, NH, Maine and RI), and on ourown Web site <http://www.boston-leos.org>, as well as the local IEEEWeb site <http://ieeeboston.org>. Inaddition we distribute at the Septemberand January LEOS meetings, and bymail to the local Universities, Colleges,and Government and CorporateResearch Labs, a one page poster of allthe upcoming fall or spring talk’s Titlesand Speakers. An electronic PDF copyof the poster also gets e-mailed to alllocal LEOS’s members, and any non-member who has added their emailaddress to the attendee’s list passedaround at all chapter meetings.Furthermore, everyone on the two e-mail lists receives an automatedremainder e-mail about each talk threedays before it occurs.

Finally, we list CNEC LEOS’s2007-08 yearly program of monthlyseminars and speakers, 3 of them localspeakers, 7 from out-of-state, one aLEOS Distinguished Lecturer, andone a Nobel Prize winner:

1. September: “Energy Transfer andLight Emission from Erbium-dopedSilicon Nanostructures,” Prof. LucaDal Negro, Boston University,Boston, MA.

2. October: “Semiconductor Lasers:From the Ultra-fast to the Ultra-small,” Prof. Farhan Rana, CornellUniversity, Ithaca, NY.

3. November: “Compact HighRepetition Rate Soft X-Ray Lasers: ADoorway To High Intensity CoherentSoft X-Ray Science On A Table-Top’”Prof. Jorge J. Rocca, an IEEE LEOSDistinguished Lecturer, ColoradoState University, Fort Collins, CO.

4. December: -- This meeting was can-

celed at last minute due to snowstorm– “Seeing The Invisible:Polarization Vision In Nature AndNon-Invasive Imaging AndSensing,” Prof. Nader Engheta,University of Pennsylvania,Philadelphia, PA.

5. January: “Manipulating Micro-cavities With Optical Forces AndPotentials: Toward Self-AligningSmart" Micro-cavities AndPicometer-Scale OptomechanicalControl,” Dr. Peter Rakich, MIT,Cambridge, MA.

6. February: “Advances in Real-time 3DImaging, Display, and Visualization,”Prof. Bahram Javidi, University ofConnecticut, Storrs, CT.

7. “March: “Recent Results in Mid-Infrared Quantum Cascade Lasers,”Prof. Claire F. Gmachl, PrincetonUniversity, Princeton, NJ.

8. April: “New Forms of Matter -Created, Manipulated and Observedwith Laser Light,” Prof. WolfgangKetterle, a Nobel Prize winner(Physics), MIT, Cambridge, MA.

9. May: “Applications of Slow andStopped Light , “Prof. John C.Howell, University of Rochester,Rochester, NY.

10.June: “Quantum Imaging,” Prof.Yanhua Shih, University of Maryland,Baltimore County, Baltimore, MD.

And in conclusion, we list the talksand speakers at the CNEC LEOS’sfall 2007 Plasmonics Workshop (5of them local speakers, 6 from out-of-state):

1. “Plasmonics-The Missing LinkBetween Nanoelectronics andMicrophotonics,” Prof. MarkBrongersma, Stanford University,Palo Alto, CA.

2. “Metamaterials: from Near-Infrared toVisible, “ Prof. Vladimire Shalaev,Purdue University, West Lafayette, IN.

3. “Nanostructures by the Square Yard:Large-Area Plasmonic and Negative-Index Materials,” Prof. Steve Brueck,University of New Mexico,Albuquerque, NM.

4. “Nanophotonics in Mid-IR:fromSuperlenses to Transmission Lines,”Prof. Gennady Shvets, University ofTexas, Austin, TX.

5. “Resonant Transmission of LightThrough Sub-wavelength Holes inThick Metal Films,” Prof. JorgeBravo Abad, MIT, Cambridge, MA

6. “Plasmonics at Terahertz and Mid-Infrared Frequencies,” Prof. RichardAvetitt, Boston University, MA.

7. “Optical Antennas,” Prof.Kenneth Crozier, HarvardUniversity, Cambridge, MA.

8. “Electric and Magnetic SurfacePlasmons on Metamaterials,” Prof.Willie Padilla, Boston College,Newton, MA.

9. “Coherent, Nonlinear, and UltrafastNanoplasmonics,” Prof. MarkStockman, Georgia State University,Atlanta, GA.

10.“Giant Field Enhancement andLocalization in Aperiodic PlasmonicStructures,” Prof. Luka Dal Negro,Boston University, Boston, MA.

11.“Wave-Bending Plasmonic Elements,Nanocircuits with Light, andMetaplasmonic Structures,” Prof.Nader Engheta, University ofPennsylvania, PA.

William H. Nelsonmember Technical Program Committee

LEOS CNEC Chapter

22leos03.qxd 6/16/08 1:17 PM Page 37


Conference Section

Recognition at OFC 2008

John H. Marsh, LEOS President, recognized the following LEOS members who have been elevated to the grade ofIEEE Fellow: (From left to right) Radhakrishnan Nagarajan, Patrick Iannone, Kazuo Hagimoto, Janet Jackel, JohnMarsh (LEOS President), Susumu Noda, Akihiko Kasukawa, David Welch, and Kit Lia Paul Yu.

The 2008 John Tyndall Award was presented to Robert W.Tkach, “'for pioneering breakthroughs in high-capacity trans-mission systems and networks, including the invention ofNZDF (non-zero dispersion fiber) and dispersion managementof optical fiber nonlinearities”. This award is jointly sponsoredby LEOS, OSA, and endowed by Corning Inc.

Joe C. Campbell (left) received the 2008 IEEE Photonics Award,“for seminal and sustained contributions to the development ofhigh-speed, low-noise long wavelength avalanche photodiode.The award, which is sponsored by LEOS, was presented to Dr.Campbell by Dr. Lewis M. Terman, IEEE President.

22leos03.qxd 6/16/08 1:17 PM Page 38


Conference Section (cont’d)

22leos03.qxd 6/16/08 1:17 PM Page 39


Conference Section

22leos03.qxd 6/16/08 1:17 PM Page 40



22leos03.qxd 6/16/08 1:17 PM Page 41



22leos03.qxd 6/16/08 1:17 PM Page 42



The Winter TopicalsConference in Sorr-ento, Italy 14-16January 2008, brou-ght together leadingexperts from researchand industry, provid-ing a familiar atmos-phere for discussing

new, innovative and upcoming tech-nology trends in one of the mostexciting topic of optics and photon-ics, the microstructured opticalfibers, also called Photonic CrystalFibers (PCFs) or holey fibers. PCFshave generated great interest in thescientific community as they offer thepossibility to control the propagationof light with flexibility not obtain-able with conventional optical fibers.In a PCF, light can be guided usingeither total internal reflection or aphotonic band-gap arising from somespecific periodic arrangements of theair-holes. When a defect is intro-duced into such a structure, a local-ized state is created within the band-gap, and the light guiding becomespossible along the PCF. Solid-coreand hollow-core PCFs are recognizedas promising structures to exploit dif-

ferent regimes of light propagationand different media characteristics.Due to their unique properties, PCFshave important applications in sever-al fields, such as telecommunications,metrology, spectroscopy, microscopy,astronomy, lithography, coherentoptical tomography.

The Topical Meeting on “PhotonicCrystal Fibers: Technology andApplications” gave the opportunity toexchange ideas between scientistsworking on technology and applica-tions by covering a variety of topicsrelated to PCF based devices and phe-nomena, including lasers and ampli-fiers, high nonlinearities, numericaldesign and analysis, supercontinuumgeneration, optical sensors, technolo-gies and materials, band-gap effects,wide-band transmissions, birefrin-gence. Nine regular sessions plus oneJoint Session covered the cited topics,enriched by sixteen invited papersgiven by scientists from all over theworld who provided an effectiveoverview about this rapidly growingarea, stimulating particular intereston topics like high power applica-tions, nonlinear phenomena and prop-erties of band-gap PCFs. The meeting

gave the sensation that the recent pro-gresses and the future potentials, notyet fully exploited, will for sure drivean exciting PCF development stillable to attract attention and researchefforts for years, to find ever interest-ing and new applications in manyareas of human activity.

The pleasant location on the Sorrentocoast and the proper selection of correlat-ed topics in photonics, “Chip-ScaleNonlinear Optical Devices”, “FibreOptical Parametric Amplifiers andRelated Devices” and “NonlinearOptics in Liquid Crystals”, helped ingathering a number of interested peoplefrom research and industry.

As the meeting Chair, I would liketo thank the Co-chair, Prof. KunimasaSaitoh, for the invaluable help, all theinvited speakers for the outstandingpresentations, all the participants fortheir contributions and the LEOSstaff, which helped assuring the suc-cessful organization of the meeting.

Prof. Stefano SelleriUniversity of Parma, Italy

Chair of the Winter Topicals ConferencePhotonic Crystal Fibers:

Technology and Applications

Review of 2008 LEOS Winter Topical Meetings

Stefano Selleri

“Nick” Cartoon Series by Christopher Doerr

22leos03.qxd 6/16/08 1:17 PM Page 43



22leos03.qxd 6/16/08 1:17 PM Page 44



22leos03.qxd 6/16/08 1:17 PM Page 45


Publication Section

JOURNAL OF SELECTED TOPICS INQUANTUM ELECTRONICS ON HIGHPOWER FIBER LASERSSubmission Deadline: September 1, 2008IEEE Journal of Selected Topics in Quantum Electronicsinvites manuscript submissions in the area of solidstate photonics. The purpose of this issue of JSTQE isto document recent advances in the areas of highpower fiber laser technology. Broad technical areasinclude (but are not limited to):

• Fiber design, fabrication and properties; components- design, fabrication and characterization tech-

niques for novel fibers, fiber materials and dopants- photodarkening- large-mode-area and dispersion-tailored fibers in

glass-air or all-glass geometries- distributed filtering- polarization preservation- fiber and volume Bragg gratings and other

components• Fiber lasers and amplifiers

- modeling and realization of fiber lasers and ampli-fiers for all wavelengths, linewidths, and temporalregimes

- beam propagation control and pulse shaping- thermal management- pump coupling, and all aspects of packaging and

integration• Fiber nonlinearities

- management and application of fiber nonlineari-ties (FWM, SPM, SRS, SBS, etc.)

- spatial Kerr effects

The Guest Editors for this issue are Johan Nilsson,University of Southampton – Optoelectronics ResearchCenter, Southampton, UK; Siddharth Ramachandran,OFS Labs, Somerset, NJ, USA, Thomas Shay, Air ForceResearch Labs – DELO, Kirtland AFB, Albuquerque, NM,USA, and Akira Shirakawa, University of Electro-Communications – Institute for Laser Science, Tokyo, Japan

The deadline for submission of manuscripts is September 1,2008; publication is scheduled for January/February of 2009.

Online Submission is Mandatory at: http://mc.manuscript-central.com/leos-ieee Please select the Journal of SelectedTopics Of Quantum Electronics Journal from the dropdown menu. Contributed papers should be up to eight

pages in length, and invited up to 12 pages. Beyond that,a charge of $220 per page apply. All submissions will bereviewed in accordance with the normal procedures of theJournal. You may find the Tools for Authors link useful:http://www.ieee.org/web/publications/authors/transjnl/index.html.

For inquiries please contact:JSTQE Editorial Office - Chin Tan-yan

IEEE/LEOS445 Hoes Lane

Piscataway, NJ 08854 USAPhone: 732-465-5813

Email: [email protected]

Contact [email protected] for any questions about thisissue. For all papers published in JSTQE, there are vol-untary page charges of $110.00 per page for each pageup to eight pages. Invited papers can be twelve pagesand Contributed papers should be 8 pages in lengthbefore overlength page charges of $220.00 per page arelevied. The length of each paper is estimated when it isreceived. Authors of papers that appear to be overlengthare notified and given the option to shorten the paper.Additional charges will apply if color figures arerequired.

The following supporting documents are requiredduring manuscript submission:

1) .doc manuscript (double columned, 12 pages for anInvited Paper. Contributed paper should be doublecolumned, 8 pages in length.) Bios of ALL authorsare mandatory, photos are optional. You may findthe Tools for Authors link useful:

http://www.ieee.org/web/publications/authors/tran-sjnl/index.html

2) Completed the IEEE Copyright Form. Copy andpaste the link below:http://www.ieee.org/web/publications/rights/copy-rightmain.html

3) Completed Color Agreement/decline form.If your paper is accepted, you may have it in color,online in our IEEE Explore site for free. However,if you wish for color in print, please check of the

Call for Papers

22leos03.qxd 6/16/08 1:17 PM Page 46


Publication Section (cont’d)

appropriate Agree/decline box. (Please email [email protected] to request for this form.)

4) .doc list of ALL Authors FULL Contact informationas stated below:

Last name (Family name):First name: Suffix (Dr/Prof./Ms./Mr):Affiliation:Dept:Address:Telephone:Fax:Email:Alternative Email:

JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS ON PRELIMINARY CALL: SEMICONDUCTOR LASERSSubmission Deadline: November 1, 2008We invite authors to submit manuscripts in the area ofsemiconductor lasers. The purpose of this issue ofJSTQE is to document the current state-of-the-art andbasic research in semiconductor lasers through a col-lection of original papers.

Although the issue serves as a venue for publication of fulllength journal articles expanding upon presentations atthe 2008 International Semiconductor Laser Conference,issue is open to any research topic that is relevant to thesemiconductor laser field.

The Primary Guest Editor for this issue is Prof. Luke F.Lester of the University of New Mexico, USA. The GuestCo-Editors include Prof. Thomas L. Koch of LehighUniversity, USA and Prof. Johann Peter Reithmaier ofthe Universitaet Kassel, Germany.

The deadline for submission of manuscripts is November1, 2008; publication is scheduled for May/June of 2009.

Online Submission is Mandatory at: http://mc.manu-scriptcentral.com/leos-ieee Please select the Journal ofSelected Topics Of Quantum Electronics Journal from thedrop down menu.

The following supporting documents are required duringmanuscript submission:

1) .doc manuscript (double columned, 12 pages for anInvited Paper. Contributed paper should be doublecolumned, 8 pages in length.) Bios of ALL authorsare mandatory, photos are optional. You may findthe Tools for Authors link useful:

http://www.ieee.org/web/publications/authors/transjnl/index.html

2) Completed the IEEE Copyright Form. Copy andpaste the link below:http://www.ieee.org/web/publications/rights/copy-rightmain.html

3) Completed Color Agreement/decline form.If your paper is accepted, you may have it in color,online in our IEEE Explore site for free. However,if you wish for color in print, please check of theappropriate Agree/decline box. (Please email [email protected] to request for this form.)

4) .doc list of ALL Authors FULL Contact informationas stated below:

Last name (Family name): /First name:Suffix(Dr/Prof./Ms./Mr): /Affiliation: /Dept: /Address:/Telephone: /Fax: /Email: / Alternative Email:

All submissions will be reviewed in accordancewith the normal procedures of the Journal.

For inquiries please contact:JSTQE Editorial Office - Chin Tan-yan

IEEE/LEOS445 Hoes Lane

Piscataway, NJ 08854 USAPhone: 732-465-5813

Email: [email protected]

Contact [email protected] for any questions about thisissue. For all papers published in JSTQE, there are vol-untary page charges of $110.00 per page for each page upto eight pages. Invited papers can be twelve pages andContributed papers can be 8 pages in length before over-length page charges of $220.00 per page are levied. Thelength of each paper is estimated when it is received.Authors of papers that appear to be overlength are noti-fied and given the option to shorten the paper.Additional charges will apply if color figures are required.

22leos03.qxd 6/16/08 1:17 PM Page 47


MANAGEMENTJames A. VickStaff Director, AdvertisingPhone: 212-419-7767Fax: [email protected]

Susan E. SchneidermanBusiness Development ManagerPhone: 732-562-3946Fax: [email protected]

Marion DelaneyAdvertising Sales DirectorPhone: 415-863-4717Fax: [email protected]

PRODUCT ADVERTISINGMidatlanticLisa Rinaldo Phone: 732-772-0160Fax: [email protected], NJ, PA, DE, MD, DC,KY, WV

New England/ConnecticutStan GreenfieldPhone: 203-938-2418Fax: [email protected]

New England/Eastern CanadaJody EstabrookPhone: 978-244-0192Fax: [email protected], VT, NH, MA, RICanada: Quebec, Nova Scotia,Newfoundland, Prince EdwardIsland, New Brunswick

SoutheastBill HollandPhone: 770-436-6549Fax: [email protected], NC, SC, GA, FL, AL,MS, TN

Midwest/Central Canada Dave JonesPhone: 708-442-5633Fax: 708-442-7620

[email protected] IL, IA, KS, MN, MO, NE,ND, SD, WICanada: Manitoba,Saskatchewan, Alberta

Midwest/Ontario, CanadaWill HamiltonPhone: 269-381-2156 Fax: [email protected], MI. Canada: Ontario

OhioJoe DiNardoPhone: 440-248-2456Fax: [email protected]

SouthwestSteve LoerchPhone: 847-498-4520Fax: [email protected], LA, TX, OK

So. California/Mountain StatesMarshall RubinPhone: 818-888-2407Fax: 818-888-4907 [email protected], AZ, NM, CO, UT, NV,CA 93400 & below

Northern California/Northwest/ Western CanadaLori KehoePhone: 650-458-3051Fax: [email protected], ID, MT, WY, OR, WA,CA 93401 & aboveCanada: British Columbia

Europe/Africa/Middle EastHeleen VodegelPhone: +44-1875-825-700Fax: [email protected], Africa, Middle East

Asia/Far East/Pacific RimSusan SchneidermanPhone: 732-562-3946Fax: [email protected], Far East, Pacific Rim,Australia, New Zealand

RECRUITMENT ADVERTISINGMidatlanticLisa RinaldoPhone: 732-772-0160Fax: [email protected], NJ, CT, PA, DE, MD,DC, KY, WV

New England/Eastern CanadaJohn RestchackPhone: 212-419-7578Fax: [email protected], VT, NH, MA, RICanada: Quebec, Nova Scotia,Prince Edward Island,Newfoundland, NewBrunswick

SoutheastThomas FlynnPhone: 770-645-2944Fax: [email protected], NC, SC, GA, FL, AL,MS, TN

Midwest/Texas/Central CanadaDarcy GiovingoPhone: 847-498-4520Fax: [email protected]; AR, IL, IN, IA, KS, LA, MI,MN, MO, NE, ND, SD, OH, OK, TX, WI. Canada:Ontario, Manitoba,Saskatchewan, Alberta

West Coast/Southwest/Mountain StatesTim MattesonPhone: 310-836-4064Fax: [email protected], CO, HI, NV, NM, UT,CA, AK, ID, MT, WY, OR,WA. Canada: BritishColumbia

Europe/Africa/Middle EastHeleen VodegelPhone: +44-1875-825-700Fax: [email protected], Africa, Middle East

IEEE Lasers and Electro-OpticsSociety Newsletter

Advertising Sales Offices445 Hoes Lane, Piscataway NJ 08854

www.ieee.org/ieeemediaImpact this hard-to-reach audience in their own Society

publication. For further information on product and recruitment advertising, call your local sales office.

ADVERTISER’S INDEXThe Advertiser’s Index contained in this issue iscompiled as a service to our readers and advertis-ers. The publisher is not liable for errors or omis-sions although every effort is made to ensure itsaccuracy. Be sure to let our advertisers know youfound them through the IEEE LEOS Newsletter.

Advertiser’s Index . . . . . . . . . . .Page #

R Soft . . . . . . . . . . . . . . . . . . . . . CVR2

Tempo Plastics . . . . . . . . . . . . . . . . . 3

Mathworks . . . . . . . . . . . . . . . . . . . . 5

Optiwave . . . . . . . . . . . . . . . . . . . . . . . . . 7

IEEE MDL . . . . . . . . . . . . . . . . . . . . . 9

CSI . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Corning Incorporated . . . . . . . . . . . 17

Third Millenium . . . . . . . . . . . . . . . 21

IEEE Marketing . . . . . . . . . . . . . CVR3

General Photonics . . . . . . . . . . . CVR4

LEOS Mission StatementLEOS shall advance the interests of its mem-bers and the laser, optoelectronics, and pho-tonics professional community by:• providing opportunities for informa-

tion exchange, continuing education,and professional growth;

• publishing journals, sponsoring con-ferences, and supporting local chapterand student activities;

• formally recognizing the professionalcontributions of members;

• representing the laser, optoelectronics,and photonics community and servingas its advocate within the IEEE, thebroader scientific and technical com-munity, and society at large.

LEOS Field of InterestThe Field of Interest of the Society shall belasers, optical devices, optical fibers, andassociated lightwave technology and theirapplications in systems and subsystems inwhich quantum electronic devices are keyelements. The Society is concerned with theresearch, development, design, manufac-ture, and applications of materials, devicesand systems, and with the various scientificand technological activities which con-tribute to the useful expansion of the fieldof quantum electronics and applications.

The Society shall aid in promoting closecooperation with other IEEE groups andsocieties in the form of joint publications,sponsorship of meetings, and other forms ofinformation exchange. Appropriate cooper-ative efforts will also be undertaken withnon-IEEE societies.

22leos03.qxd 6/16/08 1:17 PM Page 48

IEEE.tv is made possible by the members of IEEE

Tune in.www.ieee.org/ieeetv

BETA

Introducing

FEATURING: Conference Highlights

Author Profi les

Careers in Technology

Reports from IEEE Spectrum

Special Features

and more!

New programs every month!

Exclusive programming fortechnology professionals.

access via

887-Qa IEEEtv 7x10.indd 1 8/16/06 11:15:16

22leos03.qxd 6/16/08 1:17 PM Page 49

22leos03.qxd 6/16/08 1:17 PM Page 50

THE SOCIETY FOR PHOTONICSTechnologies (PhAST). PhAST is possibly the least prominent of these...

Documents

Transcript of THE SOCIETY FOR PHOTONICSTechnologies (PhAST). PhAST is possibly the least prominent of these...