Download - Copyright Gordon Bell & Jim Gray PC+ The PC+ Era Infinite processing, memory, and bandwidth @ zero cost. Gordon Bell Bay Area Research Center Microsoft.

Copyright Gordon Bell & Jim Gray PC+Copyright Gordon Bell & Jim Gray PC+

The PC+ EraInfinite processing, memory, and bandwidth @ zero cost.

Gordon Bell

Bay Area Research Center

Microsoft Corporation

The only thing that matters at the end of the day is, it’s a great building.


The Highly Probable Future c2025 83 items from J. Coates, Futurist, Vol. 84, 1994

8.4 B, english speaking, personally tagged & identified, prosthetic assisted and/or mutant, tense people who have access & control of their medical records

Everything will be smart, responsive to environment.– Sensing of everything… challenge for science & engineering!– Fast broadband network– Smart appliances & AI – Tele-all: shop, vote, meet, work, etc.– Robots do everything, but there may be conflict with labor…

A “managed”, physical and man-made world– Reliable weather reports– “Many natural disasters e.g. floods, earthquakes, will be mitigated,

controlled or prevented” Nobel prize to “economist” for “value of information”


PC At An Inflection Point

PCsPCsNon-PCNon-PCdevices and Internetdevices and Internet


The Dawn Of The PC-Plus Era, The Dawn Of The PC-Plus Era, Not The Post-PC Era…Not The Post-PC Era…

Consumer Consumer PCsPCsTV/AVTV/AV MobileMobile

CompanionsCompanions

Household Household ManagementManagement

CommunicationsCommunications Automation Automation & Security & Security


PCTV a.k.a. MilliBillgUsing PCs to drive large screens e.g. tv sets, Plasma Panels

Gordon BellJim Gemmell

Bay Area Research CenterMicrosoft Research

Copyright 1999 Microsoft Corporation

HomeCATV

Analog/digital cable distribution

PC broadcasts are mixed into home CATV in analog and/or MPEG digital

Ethernet Home network

Video capture

“milliBill”

Basic ideas:

1. PC records or plays thru video cable channels. 2. PC “broadcasts” art images, webcams, presentations,

videos, DVDs, etc.3. Ethernet not cable?

Settopbox

Another big bang? Internet to TV and audio: The Net, PC meet the TV


PC will prevail for the next decade as the dominant platform… to HPCC community its COTS! Moore’s Law to reduce price Lack of last mile bandwidth to move

pictures, data, and interact favors home mainframes aka PCs

Very large disks (1TB by 2005) to “store everything” personal

Screens to enhance use Home entertainment server… Office and portable requirements Etc.


My betting record: No losses … so far (>5year old bets)

TMC & MPP will not be dominant by 1995 Video On Demand will not exist by 1995 AT&T acquisition of NCR will not be successful 10K desktop-desktop will not exist by 1/2001 1 B internet users by 1/2001 or 1/2002 Cars won’t drive themselves by 2005 PCs continue with 2 digit growth through 2002


Outline Future predictions… 2020 and the world Caveat: How far out can we see? WWW just >5 years old Background: SNAP at RCI 3/95 conference, Albuquerque My own history of supercomputing… data/compute The hardware scene in 5-10 years?

– Processing and Moore’s Law– Networking– Disks

Challenges:– OSS– Communities with dbases & hs nets– ASP: workbenches– If simulation is third mode after theory, expt, what is 4th? connection with the experimental

world for data; then control… biologist workbench where work is being done.


SNAP … as given at RCI, 3/95

Scalable Network And Platforms A View of Computing in 2000+

(I missed the impact of WWW)

Gordon Bell Jim Gray

NetworkPlatform


How Will Future Computers Be Built?

Thesis: SNAP: Scalable Networks and Platforms• upsize from desktop to world-scale computer• based on a few standard components• similar to NEC’s Computers &

Communications 1983 visionBecause:

• Moore’s law: exponential progress

• Standardization & Commoditization• Stratification and competition

When: Sooner than you think!• massive standardization gives massive use • economic forces are enormous

NetworkPlatform


Performance versus time for various microprocessors

2000199819961994199219901988198619841982198019781

10

100

1000

DECPCMIPS


Volume drives simple,cost to standardplatforms

MPPs1-4 processor mP

1-20 processor mP

Distributed workstations

Clustered Computers

price for high speed

interconnect

price

performance

Stand-aloneDesk tops

PCs


Section: The economics of operating systems and databases(or why NT has the advantageover proprietary or vanitychips and UNIX dialects )


The UNIX Trap:creating the myth of “open systems”

“Standard” now means different!

VendorIX platforms have created the “downsizing” market that provides an apparent, order of magnitude cost reduction

Hardware platform vendors lock-in users with servers ofproprietary UNIX dialects and unique chipsto maintain margins for chip and UNIX development

Users hostage with client-server, database, and apps

An implicit or unconscious cartel forms that maintains the industry status quo


The UNIX Cartel and Tax:It’s not competitive andit introduces higher downstream costs 10,000 programmers @75 companies maintain dialects

R & D costs $1.4 - $2 billion

Implied selling price $10 - 14 billion for $1.4 billion, or a sales tax of 1 million UNIX units of $10,000

Cost could be reduced to $400 million for ONE UNIX,sales price for 1 million units would be $2,400 - 4,000

NT sales price is $650; OS2 needs to sell for $1.2b/6m

Furthermore:

The downstream effects on database vendors is40% R&D efficiency causingan implied database tax of 2.5x the sales price!

The downstream effects on apps vendors is similar

xx


Section: SNAP Architecture----------


ComputingSNAPbuilt entirelyfrom PCs

Local & global data commworld

ATM† & LocalArea Networksfor: terminal,

PC, workstation,& servers

Centralized& departmental

uni- & mP servers(UNIX & NT)

Legacymainframes &

minicomputersservers & terms

Wide-areaglobal

ATM network

Legacymainframe &

minicomputerservers & terminals

Centralized& departmental

servers buit fromPCs

scalable computers

built from PCs

TC=TV+PChome ...

(CATV or ATM or satellite)

???

Portables

A space, time (bandwidth), & generation scalable environment

Person servers (PCs)

Person servers (PCs)

MobileNets


GB with NT, Compaq, & HP cluster

In a decade we can/will have: more powerful personal computers

– processing 10-100x– 4x resolution (2K x 2K) displays to impact paper– Large, wall-sized and watch-sized displays– low cost, storage of one terabyte for personal use

adequate networking? PCs now operate at 1 Gbps– ubiquitous access = today’s fast LANs– Competitive wireless networking

One chip, networked platforms e.g. light bulbs, cameras everywhere, etc. managed by PCs!

Some well-defined platforms that compete with the PC for mind (time) and market sharewatch, pocket, body implant, home

Inevitable, continued cyberization… the challenge… interfacing platforms and people.


High Performance Computing

A 60+ year view


Star Bridge


Linux super howls


Dead Supercomputer Society


Dead Supercomputer Society ACRI Alliant American Supercomputer Ametek Applied Dynamics Astronautics BBN CDC Convex Cray Computer Cray Research Culler-Harris Culler Scientific Cydrome Dana/Ardent/Stellar/Stardent Denelcor Elexsi ETA Systems Evans and Sutherland Computer Floating Point Systems Galaxy YH-1

Goodyear Aerospace MPP Gould NPL Guiltech Intel Scientific Computers International Parallel Machines Kendall Square Research Key Computer Laboratories MasPar Meiko Multiflow Myrias Numerix Prisma Tera Thinking Machines Saxpy Scientific Computer Systems (SCS) Soviet Supercomputers Supertek Supercomputer Systems Suprenum Vitesse Electronics


Steve Squires & Cray


0.0001

0.001

0.01

0.1

1

10

100

1000

1985 1990 1995 2000 2005 2010

Bell Prize and Future Peak Tflops (t)

Petaflops study target

NEC

XMP NCube

CM2

*IBM


Top 10 tpc-c

Top two Compaq systems are:Top two Compaq systems are:1.1 & 1.5X faster than IBM SPs;1.1 & 1.5X faster than IBM SPs;1/3 price of IBM1/3 price of IBM1/5 price of SUN1/5 price of SUN


High Performance Computing

Supers we knew are Japanese… scalability & COTS in… but you have to roll your

own else pay the Unix & proprietary taxes Beowulf is $14K/TB ( 6 x 4 x 40 GB) IBM 4000R 1 rack: 2x42 500Mhz processors, 84

GB, 84 disks (3TB @36GB/disk)$420K … still cheaper than the “big buys”

$10-20K/node for special purpose vs $2K for a MAC

EMC, IBM at $1 million/TB; vs $14K


High performance architectures timeline

1950 . 1960 . 1970 . 1980 . 1990 . 2000Vtubes Trans. MSI(mini) Micro RISC nMicr

“IBM PC”Processor overlap, lookahead “killer micros”

Cray era 6600 7600 Cray1 X Y C TFunc Pipe Vector-----SMP---------------->

SMP mainframes---> “multis”----------->DSM?? Mmax.KSR SGI---->

Clusters TandmVAX IBM UNIX->

MPP if n>1000 Ncube Intel IBM->

Local NOW and Global Networks n>10,000 Grid


High performance architectures timeline


“IBM PC”Sequential programming---->------------------------------

(single execution stream e.g. Fortran) Processor overlap, lookahead “killer micros”

Cray era 6600 7600 Cray1 X Y C TFunc Pipe Vector-----SMP---------------->

SMP mainframes---> “multis”----------->DSM?? Mmax.KSR DASHSGI--->

<SIMD Vector--//--------------- Parallelization---

-----------------THE NEW BEGINNING-----------------------Parallel programs aka Cluster Computing <---------------multicomputers <--MPP era------Clusters TandmVAX IBM UNIX->MPP if n>1000 Ncube Intel IBM->Local NOW Beowlf

and Global Networks n>10,000 Grid


High performance architecture/program timeline


Sequential programming---->------------------------------(single execution stream)<SIMD Vector--//--------------- Parallelization---

Parallel programs aka Cluster Computing <---------------multicomputers <--MPP era------ultracomputers 10X in size & price! 10x MPP

“in situ” resources 100x in //sm NOW VLSCCgeographically dispersed Grid


Computer types

NetwrkedSupers…

GRIDLegionCondor Beowulf NT clusters

VPPuni

T3E SP2(mP) NOW

NEC mP

SGI DSM clusters &SGI DSM

NEC super Cray X…T(all mPv)

MainframesMultis

WSs PCs

-------- Connectivity--------

WAN/LAN SAN DSM SM

mic

ros

v

ecto

r

Clusters


Technical computer types:Pick of: 4 nodes, 2-3 interconnects

Fujitsu Hitachi

IBM ?PC? SGI cluster Beow/NT

NEC

SGI DSM T3 HP?

NEC super Cray ???FujitsuHitachiHP IBM

Intel SUNplain old

PCs

SAN DSM SMP

mic

ros

v

ecto

r


Technical computer types

NetwrkedSupers…

GRID

LegionCondor Beowulf

VPPuni

SP2(mP) NOW

NEC mP

T series



MainframesMultis

WSs PCs

WAN/LAN SAN DSM SM

mic

ros

v

ecto

r

OldWorld( one

programstream)

New world: Clustered

Computing(multiple program

streams)


Technical computer types

NetwrkedSupers…

GRID

LegionCondor Beowulf

VPPuni

SP2(mP) NOW

NEC mP

T series



MainframesMultis

WSs PCs

WAN/LAN SAN DSM SM

mic

ros

v

ecto

r

VectorizeParallellelizeMPI, Linda, PVM,

Cactus, ???distributed function

Computing Parallellelize


Gaussian Parallelism


Beyond Moore’s Law …>10 yrs Just FCB (faster, cheaper, better)…

COTS will soon mean consumer off the shelf Moore’s Law and technology progress likely to continue for

another decade for: processing & memory,

storage, LANs, & WANs are really evolving System-on-a chip of interesting sizes will emerge to create 0 cost

systems No DNA, molecular, or quantum computers, or new stores Any displacement technology is unlikely

… Carver Mead’s Law c1980A technology takes 11 years to get established

On the other hand, we are on Internet time!


We get more of everything


Computer ops/sec x word length / $

y = 1E-248e0.2918x

1.E-06

1.E-03

1.E+00

1.E+03

1.E+06

1.E+09

1880 1900 1920 1940 1960 1980 2000

.=1.565^(t-1959.4)

doubles every 7.5

doubles every 2.3

doubles every 1.0


0.01

0.1

1

10

100

1000

10000

1986

1988

1990

1992

1994

1996P

erf

orm

an

ce in

Mfl

op

/s

Micros

Supers

8087 802876881

80387

R2000

i860

RS6000/540Alpha

RS6000/590Alpha

Cray 1S

Cray X-MP

Cray 2 Cray Y-MP Cray C90Cray T90

1998

Growth of microprocessor performance

1980

1982


Albert Yu predictions ‘96

When 2000 2006

Clock (MHz) 900 4000

MTransistors 40 350

Mops 2400 20,000

Die (sq. in.) 1.1 1.4


Processor Limit: DRAM GapµProc60%/yr..

DRAM7%/yr..

1

10

100

10001980

1981

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

DRAM

CPU

1982

Processor-MemoryPerformance Gap:(grows 50% / year)

Per

form

ance

• Alpha 21264 full cache miss / instructions executed: 180 ns/1.7 ns =108 clks x 4 or 432 instructions• Caches in Pentium Pro: 64% area, 88% transistors*Taken from Patterson-Keeton Talk to SigMod

“Moore’s Law”


Sony Playstation export limiits

System-on-a-chip alternatives

FPGA Sea of un-committed gate arrays

Xylinx, Altera

Compile a system

Unique processor for every app

Tensillica

Systolic | array

Many pipelined or parallel processors

DSP | VLIW

Special purpose processors

TI

Pc & Mp.

ASICS

Gen. Purpose cores. Specialized by I/O, etc.

Intel, Lucent, IBM

Universal Micro

Multiprocessor array, programmable I/o

Cradle

Cradle: Universal Microsystemtrading Verilog & hardware for C/C++

Single part for all apps Programming @ run time via FPGA & ROM 5 quad mPs at 3 Gflops/quad = 15 Glops Single shared memory space, caches Programmable periphery including:

1 GB/s; 2.5 GipsPCI, 100 baseT, firewire

$4 per flops; 150 mW/Gflops

UMS : VLSI = microprocessor : special systemsSoftware : Hardware

MSP

MSP

MSP

M EM O R Y

MSP

MSP

MSP

MSP

M EM O R Y

MSP

MSP

MSP

MSP

M EM O R Y

C LO C KS,D EBU G

MSP

MSP

MSP

MSP

M EM O R YD R AMC O N TR O L

MSP

D R AM

PR O G I/O PR O G I/O PR

OG

I/O

PR

OG

I/O

PR

OG

I/O

PROG I/OPROG I/OPROG I/OPROG I/O

PR

OG

I/OP

RO

G I/O

PR

OG

I/O

N VM EM

UMS Architecture

Memory bandwidth scales with processing Scalable processing, software, I/O Each app runs on its own pool of processors Enables durable, portable intellectual property


Free 32 bit processor core

Linus’s Law: Linux everywhere

Software is or should be free All source code is “open” Everyone is a tester Everything proceeds a lot faster when

everyone works on one code Anyone can support and market the code for

any price Zero cost software attracts users! All the developers write lots of code


ISTORE Hardware Vision

System-on-a-chip enables computer, memory, without significantly increasing size of disk

5-7 year target:MicroDrive:1.7” x 1.4” x 0.2”

2006: ?1999: 340 MB, 5400 RPM,

5 MB/s, 15 ms seek2006: 9 GB, 50 MB/s ? (1.6X/yr capacity, 1.4X/yr BW)

Integrated IRAM processor2x height

Connected via crossbar switchgrowing like Moore’s law

16 Mbytes; ; 1.6 Gflops; 6.4 Gops10,000+ nodes in one rack! 100/board = 1 TB; 0.16 Tf


The Disk Farm? or a System On a Card?

The 500GB disc cardAn array of discsCan be used as 100 discs 1 striped disc 50 FT discs ....etcLOTS of accesses/second of bandwidth

A few disks are replaced by 10s of Gbytes of RAM and a processor to run Apps!!

14"


Nanochip.com

8

Trends: promises NEMS (Nano Electro Mechanical Systems)(http://www.nanochip.com/) also Cornell, IBM, CMU,…

• 250 Gbpsi by using tunneling electronic microscope

• Disk replacement

• Capacity: 180 GB now, 1.4 TB in 2 years

• Transfer rate: 100 MB/sec R&W

• Latency: 0.5msec

• Power: 23W active, .05W Standby

• 10k$/TB now, 2k$/TB in 2002


Disk vs TapeAt 10K$/TB disks are competitive with nearline tape.

Disk– 40 GB– 20 MBps– 5 ms seek time– 3 ms rotate latency– 7$/GB for drive

3$/GB for ctlrs/cabinet– 4 TB/rack

– 1 hour scan

Tape– 40 GB– 10 MBps– 10 sec pick time– 30-120 second seek time– 2$/GB for media

8$/GB for drive+library– 10 TB/rack

– 1 week scan

The price advantage of tape is narrowing, and the performance advantage of disk is growing

GuestimatesCern: 200 TB3480 tapes2 col = 50GBRack = 1 TB=20 drives

1988 Federal Plan for Internet

Telnet & FTP

EMAIL

WWW Audio Video

Voice!Voice!

StandardsStandards

Increase Capacity(circuits & bw)

Lower response time

Create newservice

Increased Demand

The virtuous cycle of bandwidth supply and demand


744Mbps over 5000 km to transmit 14 GB

~ 4e15 bit meters per second

4 Peta Bmps (“peta bumps”)Single Stream tcp/ip throughput

Information Sciences InstituteMicrosoft

QWestUniversity of Washington

Pacific Northwest GigapopHSCC (high speed connectivity

consortium)DARPA


Map of Gray Bell Prize resultsRedmond/Seattle, WA

San Francisco, CA

New York

Arlington, VA

5626 km10 hops

single-thread single-stream tcp/ip single-thread single-stream tcp/ip via 7 hopsvia 7 hops desktop-to-desktop …Win 2K desktop-to-desktop …Win 2K

out of the box performance*out of the box performance*


1 GBps1 GBps

Ubiquitous 10 GBps SANs in 5 years

1Gbps Ethernet are reality now.– Also FiberChannel ,MyriNet, GigaNet,

ServerNet,, ATM,…

10 Gbps x4 WDM deployed now (OC192)

– 3 Tbps WDM working in lab In 5 years, expect 10x,

wow!!

5 MBps20 MBps

40 MBps

80 MBps

120 MBps120 MBps(1Gbps)(1Gbps)


0

50

100

150

200

250

100Mbps Gbps SAN

Transmitreceivercpusender cpu

Time µs toSend 1KB

The Promise of SAN/VIA:10x in 2 years http://www.ViArch.org/

Yesterday: – 10 MBps (100 Mbps Ethernet)

– ~20 MBps tcp/ip saturates 2 cpus

– round-trip latency ~250 µs

Now– Wires are 10x faster

Myrinet, Gbps Ethernet, ServerNet,…

– Fast user-level communication

- tcp/ip ~ 100 MBps 10% cpu- round-trip latency is 15 us

1.6 Gbps demoed on a WAN


How much does wire-time cost? $/Mbyte?Odlyzko, 1998 & Jim Gray

Seat cost$/3y

BandwidthB/s $/MB Time

GBpsE 2000 1.00E+08 2.E-07 0.010100MbpsE 700 1.00E+07 7.E-07 0.100OC12 12960000 5.00E+07 3.E-03 0.020OC3 3132000 3.00E+06 1.E-02 0.333T1 28800 1.00E+05 3.E-03 10.000DSL 2300 4.00E+04 6.E-04 25.000POTS 1180 5.00E+03 2.E-03 200.000

Cost

($) Time Gbps Ethernet .2µ 10 ms 100 Mbps Ethernet .3µ 100 ms OC12 (650 Mbps) .003 20 ms DSL .0006 25 sec POTs .002 200 sec Wireless .80 500 sec


Modern scalable switches … are also supercomputers

Scale from <1 to 120 Tbps 1 Gbps ethernet switches scale to

10s of Gbps, scaling upward SP2 scales from 1.2


So where are the challenges?

Continued development based on clusters… Scalar processors need to compete with vectors. The U.S. has cast its lot with COTS!

WWW is here. Now exploit it in every respect.– Exploit OSS!

Grid Application Service Providers for scientific

and technical apps– Biologist and chemist workbenches are prototypes– Labscape @ Cell laboratory, U. of WA– Sloan sky survey


Labscape1st, 2nd, 3rd, or New Paradigm for science?


Labscape


Labscape sensors

Location tracking of people/samples– multiple resolutions – passive and active tags

Manual tasks (e.g., use of reagents, tools)

Audio/video records, vision and indexing Networked instruments (e.g., pipettes,

refrigerators, etc.)


What am I willing to predict? Processing can be anywhere…

– Maui… in the winter. BW is the limiter!– Japan… if supers are so super, otherwise use PCs– In the disks– Application Service Providers: separation of our data from

ourselves and businesses The GRID e.g. biologist & chemist workbenches iff the

IP doesn’t get in way Collaboration ala astrophysics (high energy physics,

math, earth sci. and any pure science if pure science continues!)

OSS is the big bang for supercomputing??


The End