Tampere 01 Course 4
-
Upload
prasannakumar7 -
Category
Documents
-
view
213 -
download
0
Transcript of Tampere 01 Course 4
-
7/30/2019 Tampere 01 Course 4
1/81
Enabling Technologies for
Reconfigurable Computing
Enabling Technologies for
Reconfigurable Computing
Part 4:
FPGAs: recent developments
Wednesday, November 21, 16.00 17.30 hrs.
Reiner Hartenstein
University ofKaiserslautern
November 21, 2001, Tampere, Finland
-
7/30/2019 Tampere 01 Course 4
2/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
2
Schedule
time slot
08.30 10.00 Reconfigurable Computing (RC)
10.00 10.30 coffee break
10.30 12.00 Stream-based Computing for RC
12.00 14.00 lunch break
14.00 15.30 Resources for RC
15.30 16.00 coffee break16.00 17.30 FPGAs: recent developments
-
7/30/2019 Tampere 01 Course 4
3/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
3
>> Configware Market
Configware Market FPGA Market
Embedded Systems (Co-Design)
Hardwired IP Cores on Board
Run-Time Reconfiguration (RTR)
Rapid Prototyping & ASIC Emulation
Evolvable Hardware (EH)
Academic Expertise
ASICs dead
Soft CPU
HLLs
Problems to be solved
-
7/30/2019 Tampere 01 Course 4
4/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
4
Configware heading for mainstream
Configware market taking off for mainstream FPGA-based designs more complex, even SoC No design productivity and quality without good
configware libraries (soft IP cores) from various
application areas. Growing no. of independent configware houses
(soft IP core vendors) and design services AllianceCORE & Reference Design Alliance
Currently the top FPGA vendors are the keyinnovators and meet most configware demand.
-
7/30/2019 Tampere 01 Course 4
5/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
5
bleeding edge designs
Infinite amount of gates not yet available on a chip 3 mio gates (10 mio in 2003 ?) far away from "infinite"
Bleeding edge designs only with sophisticated EDA tools
Excessive optimization needed Hardware epertise is inevitable for the designer.
improve and simplify the design flow the user
provide rich configware libraries of soft IP cores,
control appl., networking, wireless telecommunication, datacommunication, embedded and consumer markets.
-
7/30/2019 Tampere 01 Course 4
6/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
6
Configware (soft IP Products)
For libraries, creation and reuse of configware To search for IPs see: List of all available IP
The AllianceCORE program is a cooperationbetween Xilinx and third-party core developers
The Xilinx Reference Design Alliance Program
The Xilinx University Program
LogiCORE soft IP with LogiCORE PCI Interface. Consultants
-
7/30/2019 Tampere 01 Course 4
7/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
7
EDA as the Key Enabler (major EDA vendors)
Select EDA quality / productivity, not FPGA architectures EDA often has massive software quality problems
Customer: highest priority EDA center of excellence collecting EDA expertise and EDA user experience
to assemble best possible tool environments for optimum support design teams
to cope with interoperability problems
to keep track with the EDA scene as a rapidly moving target
being fabless, FPGA vendors spend most qualified manpowerin development of EDA, IP cores, applications , support
Xilinx and Altera are morphing into EDA companies.
-
7/30/2019 Tampere 01 Course 4
8/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
8
OS for FPGAs
separate EDA software market, comparable tothe compiler / OS market in computers,
Cadence, Mentor, Synopsys just jumped in.
< 5% Xilinx / Altera income from EDA SW
Changing EDA Tools Market Major configwareEDA vendors Altera Cadence
Mentor Graphics Synopsys Xilinx
-
7/30/2019 Tampere 01 Course 4
9/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
9
EDA Software for Xilinx
Full design flow from Cadence, Mentor, & SynopsysXilinx Software AllianceEDA Program:
Alliance Series Development System.
Foundation Series Development Systems.
Xilinx Foundation Series ISE (Integrated Synthesis Environment)
free WebPOWERED SW w. WebFitter & WebPACK-ISE
StateCAD XE and HDL Bencher
Foundation Base ExpressFoundation ISE Base Express
-
7/30/2019 Tampere 01 Course 4
10/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
10
Foundation ISE Base Express
ModelSim Xilinx Edition(ModelSim XE)
Forge Compiler
Modular Design Chipscope ILA
The Xilinx SystemGenerator
XPower
JBits SDK
The Xilinx XtremeDSPInitiative
MathWorks / XilinxAlliance
System Generator
Wind River / Xilinxalliance
-
7/30/2019 Tampere 01 Course 4
11/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
11
Altera EDA
Altera was founded in June 1983
EDA: synthesis, place & route, and, verification
Quartus II: APEX, Excalibur, Mercury, FLEX 6000 families
MAX+PLUS II: FLEX, ACEX & MAX families
Flow with Quartus II: Mentor Graphics, Synopsys, Synplicitydeliver a design design software to support Altera SOPC solutions.
Mentor: only EDA vendor w. complete design environment f. APEXII incl. IP, design capture, simulation, synthesis, and h/s co-verification
Configware: Altera offers over a hundred IP cores
Third party IP core design services and consultants
-
7/30/2019 Tampere 01 Course 4
12/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
12
Cadence
FPGA Designer: top-down FPGA design system, high-level mapping, architecture-specific optimization,
Verilog,VHDL, schematic-level design entry.
Verilog, VHDL to Synergy (logic synthesis) and FPGA Designer
FPGAs simulated by themselves using Cadence's Verilog-XL or Leapfrog VHDL simulators and
simulated w. rest of the system design w. Logic
Workbench board/system verification envment. Libraries for the leading FPGA manufacturers.
-
7/30/2019 Tampere 01 Course 4
13/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
13
Mentor Graphics
System Design and Verification.
PCB design and analysis:
IC Design and Verification
shifts ASIC design flow to FPGAs (Altera, Xilinx) by FPGA Advantage with IP support by ModuleWare, Xilinx CORE Generator Altera MegaWizard integration,
-
7/30/2019 Tampere 01 Course 4
14/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
14
Synopsys
FPGA Compiler II
Version of ASIC Design Compiler Ultra
Block Level Incremental Synthesis (BLIS)
ASIC FPGA migration
Actel, Altera, Atmel, Cypress, Lattice,Lucent, Quicklogic, Triscend, Xilinx
-
7/30/2019 Tampere 01 Course 4
15/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
15
>> FPGA Market
Configware Market
FPGA Market
Embedded Systems (Co-Design)
Hardwired IP Cores on Board
Run-Time Reconfiguration (RTR)
Rapid Prototyping & ASIC Emulation Evolvable Hardware (EH)
Academic Expertise
ASICs dead
Soft CPU
HLLs
Problems to be solved
-
7/30/2019 Tampere 01 Course 4
16/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
16
Top 4 PLD Manufacturers 2000
Xilinx42%
Altera
37%
Lattice15%
Actel6%
Top 4 PLD Manufacturers 2000
$3.7 Bio
-
7/30/2019 Tampere 01 Course 4
17/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
17
FPGA market 1998 / 1999
3832Atmel84030Quicklogic7
4341Cypress6
120100Lucent5
172154Actel4 410206Lattice3
837654Altera2
899629Xilinx119991998
global sales (mio $)
1999 rankSource:IC Insights Inc.
Meanwhile,
Xilinx acquiredPhilips' MOS PLDbusiness,
Latticepurchased Vantis.
.
-
7/30/2019 Tampere 01 Course 4
18/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
18
.... into every application
[Dataquest] PLD market > $7 billion by 2003.
fastest growing segment of semiconductormarket.
IP reuse and "pre-fabricated" components forthe efficiency of design and use for PLDs
FPGAs are going into every type of application.
-
7/30/2019 Tampere 01 Course 4
19/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
19
.... going into every type of application[Gordon Bell]
-
7/30/2019 Tampere 01 Course 4
20/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
20
Xilinx
fabless FPGA semi vendor, San Jose, Ca, founded 1984
key patents on FPGAs (expiring in a few years)
Fortune 2001: No. 14 Best Company to work for in (intel:no. 42, hp no. 64, TI no. 65).
DARPA grant (Nov99) to develop Jbits API tools forinternet reconfigurable / upgradable logic (w. VT)
Less brilliant early/mid 90ies (president Curt Wozniak):1995 market share from 84% down to 62% [Dataquest]
As designs get larger, Xilinx losed its advantage (bugfixesdid not require to burn new chips)
meanwhile, weeks of expensive debug time needed
-
7/30/2019 Tampere 01 Course 4
21/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
21
Xilinx Flexware
Virtex, Virtex-II, first w. 1 mio system gates.
Virtex-E series > 3 mio system gates. Virtex-EM on a copper process & addit. on chip memory f. network switch appl. The Virtex XCV3200E > 3 million gates, 0.15-micron technology,
Spartan, Spartan-XL, Spartan-II for low-cost, high volume applications as ASIC replacements Multiple I/O standards, on-chip block RAM, digital delay lock loops eliminate phase lock loops, FIFOs, I/O xlators , system bus drivers
XC4000XV, XC4000XL/XLA, CPLD: low-cost families rapid development, longer system life, robust field upgradability support In-System Programming (ISP), in-board debugging, test during manufacturing, field upgrades, full JTAG compliant interface
CoolRunner: low power, high speed/density, standby mode. Military & Aerospace: QPRO high-reliability QML certified Configuration Storage Devices
-
7/30/2019 Tampere 01 Course 4
22/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
22
Altera Flexware
Newer families: APEX 20KE, APEX 20KC, APEX II, MAX7000B, ACEX 1K, Excalibur, Mercury families.
Apex EP20K1500E (0.18-), up to 2.4 mio system gates,
APEX II (all-copper 0.13-) f. data path applications, supportsmany I/O standards. 1-Gbps True-LVDS performance
wQ2001, an ARM-based Excalibur device
Altera mainstream: MAX 7000A, 3000A; FLEX 6000,10KA, 10KE; APEX 20K families.
Mature and other : Classic, MAX 7000, 7000S, 9000;FLEX 8000, 10K families.
-
7/30/2019 Tampere 01 Course 4
23/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
23
Triscend CSoC
Digital Filter Display Interface
Viterbi A/D Interface
CSI Socket
ARM
Configurable system logic
Configurable System Interconnect (CSI) Bus
Other System ResourcesMemory
[Kean]
-
7/30/2019 Tampere 01 Course 4
24/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
24
>> Embedded Systems (Co-Design)
Configware Market
FPGA Market
Embedded Systems (Co-Design)
Hardwired IP Cores on Board
Run-Time Reconfiguration (RTR)
Rapid Prototyping & ASIC Emulation Evolvable Hardware (EH)
Academic Expertise
ASICs dead
Soft CPU
HLLs
Problems to be solved
-
7/30/2019 Tampere 01 Course 4
25/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
25
Goal: away from complex design flow
Placeand
RouteNetlist
Schematics/
HDL Netlister
Bitstream
CompilerHLL
[ la S. Guccione]
-
7/30/2019 Tampere 01 Course 4
26/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
26
Overcome traditional separate design flow
UserCode
Compiler Executable
Netlister NetlistPlace
and
Route..
Bitstream
Schematics/
HDL
HLL Compiler
[ la S. Guccione]
-
7/30/2019 Tampere 01 Course 4
27/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
27
Overcome traditional co-processing designseparate flow -> JBitsDesign Flow
User
Java
Code
Java
Compiler
JBits
API
Executable
UserCode
Compiler Executable
Netlister Netlist
Place
andRoute..
Bitstream
Schematics/
HDL
[ la S. Guccione]
E b dd d h d CPU &
-
7/30/2019 Tampere 01 Course 4
28/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
28
Embedded hardw. CPU & memory coreson chip.
HLL Compiler
CPUcore
FPGA core
Memorycore
HLL Compiler
[ la S. Guccione]
d l d l
-
7/30/2019 Tampere 01 Course 4
29/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
29
new directions in application development
new directions in application development. aut. partitioning compilers: designer productivity
like CoDe-X (Jrgen Becker, Univ. of Karlsruhe),
supports Run-Time Reconfiguration (RTR), a keyenabler of error handling and fault correction bypartial re-routing the FPGA at run time, as well asremote patching for upgrading, remote debugging,
and remote repair by reconfiguration - even overthe internet.
R Ti R fi ti (RTR)
-
7/30/2019 Tampere 01 Course 4
30/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
30
>> Run-Time Reconfiguration (RTR)
Configware Market
FPGA Market
Embedded Systems (Co-Design)
Hardwired IP Cores on Board
Run-Time Reconfiguration (RTR)
Rapid Prototyping & ASIC Emulation Evolvable Hardware (EH)
Academic Expertise
ASICs dead
Soft CPU
HLLs
Problems to be solved
CPU f fi ti t
-
7/30/2019 Tampere 01 Course 4
31/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
31
CPU use for configuration management
on-board microprocessor CPU is availableanyhow - even along with a little RTOS
use this CPU for configuration management
CompilerHLL
RTR System Design
h d CPU & hi
-
7/30/2019 Tampere 01 Course 4
32/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
32
hard CPU & memory core on same chip
CPUcore
FPGA core
MemorycoreCompilerHLL
CompilerHLL
RTR System Design
C nver in fact rs f r RTR
-
7/30/2019 Tampere 01 Course 4
33/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
33
Converging factors for RTR
User
JavaCode
Java
Compiler
JBits
API
Executable
Converging factors make RTR based system design viable 1) million gate FPGA devices and co-processing withstandard microprocessors are commonplace
direct implementation of complex algorithms in FPGAs.This alone has alreadyrevolutionized FPGA design.
2) new tools like Xilinx Jbitssoftware tool suite directlysupport coprocessing and RTR.
RTR
-
7/30/2019 Tampere 01 Course 4
34/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
34
RTR
divides application into a series of sequentially executed stages, each
implemented as a separate execution module. Partial RTR partitions these stages into finer-grain sub-modules to be
swapped in as needed. Without RTR, all conf. platforms just ASIC emulators. needs a new kind of application development environments.
directly support development and debugging of RTR appl. essential for the advancement of configurable computing will also heavily influence the future system organization Xilinx, VT, BYU work on run-time kernels, run-time support, RTR
debugging tools and other associated tools. smaller, faster circuits, simplified hardware interfacing, fewer IOBs;
smaller, cheaper packages, simplified software interfaces.
Run time Mapping
-
7/30/2019 Tampere 01 Course 4
35/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
35
Run-time Mapping
run-time reconfigurable are: Xilinx VIRTEX FPGA familyRAs being part of Chameleon CS2000 series systemsUsing such devices changes many of the basic assumptionsin the HW/SW co-design process:
host/RL interaction is dynamic, needs a tiny OS like eBIOS,also to organize RL reconfiguration under host control
typical goal is minimization of reconfiguration latency(especially important in communication processors), to hideconfiguration loading latency, and,
Scheduling to find best schedule for eBIOS calls (C~side).
>> Rapid Prototyping & ASIC Emulation
-
7/30/2019 Tampere 01 Course 4
36/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
36
>> Rapid Prototyping & ASIC Emulation
Configware Market
FPGA Market
Embedded Systems (Co-Design)
Hardwired IP Cores on Board
Run-Time Reconfiguration (RTR)
Rapid Prototyping & ASIC Emulation Evolvable Hardware (EH)
Academic Expertise
ASICs dead
Soft CPU
HLLs
Problems to be solved
ASIC emulation: a new business model ?
-
7/30/2019 Tampere 01 Course 4
37/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
37
ASIC emulation: a new business model ?
ASIC emulation / Rapid Prototyping: to replace simulation
Quickturn (Cadence), IKOS (Synopsys), Celaro (Mentor)
from rack to board to chip (from other vendors, e. g.
Virtex and VirtexE family (emulate up to 3 million gates)
Easy configuration using SmartMedia FLASH cards
ASIC emulators will become obsolete within years
By RTR: in-circuit execution debugging instead of emulation
>> Evolvable Hardware (EH)
-
7/30/2019 Tampere 01 Course 4
38/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
38
>> Evolvable Hardware (EH)
Configware Market
FPGA Market
Embedded Systems (Co-Design)
Hardwired IP Cores on Board
Run-Time Reconfiguration (RTR)
Rapid Prototyping & ASIC Emulation Evolvable Hardware (EH)
Academic Expertise
ASICs dead
Soft CPU
HLLs
Problems to be solved
EH EM
-
7/30/2019 Tampere 01 Course 4
39/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
39
EH, EM, ...
"Evolvable Hardware" (EH), "Evolutionary Methods" (EM),
digital DANN, "Darwinistic Methods", and biologicallyinspired electronic systems
new research area, also a new application area of FPGAs
revival of cybernetics or bionics: stimulated by technology
evolutionary and DNA metaphor create awareness
EM sucks, also thru mushrooming funds in the EU, inJapan, Korea, and the USA
EM-related international conference series are in theirstormy visionary phase, like EH, ICES, EuroGP, GP, CEC,GECCO, EvoWorkshops, MAPLD, ICGA
EH EM
-
7/30/2019 Tampere 01 Course 4
40/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
40
EH, EM, ...
Shake-out phenomena expected, like in the past withArtificial Intelligence
should be considered as a specialized EDA scene,focusing on theoretical issues.
Genetic algorithms suck - often replacable by moreefficient ones from EDA
It is recommendable to set-up an interwoven competencein both scenes, EM scene and the highly commercialized
EDA sceneEH should be done by EDA people, rather than EM freaks.
>> Academic Expertise
-
7/30/2019 Tampere 01 Course 4
41/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
41
>> Academic Expertise
Configware Market
FPGA Market
Embedded Systems (Co-Design)
Hardwired IP Cores on Board
Run-Time Reconfiguration (RTR)
Rapid Prototyping & ASIC Emulation Evolvable Hardware (EH)
Academic Expertise
ASICs dead
Soft CPU
HLLs
Problems to be solved
BRASS (1)
-
7/30/2019 Tampere 01 Course 4
42/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
42
BRASS (1)
UC Berkeley, the BRASS group: Prof. Dr. John Wawrzynek
The Pleiades Project, Prof. Jan Rabaey, ultra-low power highperformance multimedia computing through reconfigurationof heterogeneous system modules, reducing energy byoverhead elimination, programmability at just right
granularity, parallellism, pipelining, dynamic voltage scaling. Garp integrates processor and FPGA; dev. in parallel w.compiler - software compile techniques (VLIW SWpipelining): simple pipelining schema f. broad class of loops.
SCORE, a stream-based computation model - a unifyingcomputational model. Fast Mapping for Datapaths: by a treeparsing compiler tool for datapath module mapping
BRASS (2)
-
7/30/2019 Tampere 01 Course 4
43/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
43
BRASS (2)
HSRA. new FPGA (& related tools) supports pipelining, w.retiming capable CLB architecture, implemented in a 0.4umDRAM process supporting 250MHz operation
OOCG. Object Oriented Circuit-Generators in Java
MESCAL (GSRC), the goal is: to provide a programmer'smodel and software development environment for efficientimplementation of an interesting set of applications onto afamily of fully-programmable architectures /
microarchitectures.
Berkeley claiming (1)
-
7/30/2019 Tampere 01 Course 4
44/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
44
Berkeley claiming (1)
SCORE, a stream-based computation model: the BRASS group claims
having solved the problem of primary impediment to wide-spreadreconfigurable computing, by a unifying computational model.
Remark: clean stream-based model introduced ~1980: Systolic Array
1995: Rainer Kress. Introduces reconfigurable stream-based model
Fast Mapping for Datapaths (SCORE): BRASS claims havingintroduced 1998 the first tree-parsing compiler tool for datapathmodule mapping ." Further, it is the first work to integrate
simultaneous placement with module mapping in a way that preserveslinear time complexity."
Berkeley claiming (2)
-
7/30/2019 Tampere 01 Course 4
45/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
45
Berkeley claiming (2)
Remark: The DPSS (Data Path Synthesis System) usingtree covering simultanous datapath placement and routinghas been published in 1995 by Rainer Kress
Chip-in-a-Da2 Bee Project. Prof. Dr. Bob Brodersonsradical rethink of the ASIC design flow aimed atshortening design time, relying on stream-based DPUarrays. [published in 2000]
Remark: the KressArray, a scalable rDPU array [1995] isstream-based
.... Stream Processors -MSP-3
-
7/30/2019 Tampere 01 Course 4
46/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
46
.... m M
3rd Workshop on Media and Stream Processors (MSP-3)
http://www.pdcl.eng.wayne.edu/msp01 in conj. w. 34th Intl Symp. on Microarchitecture (MICRO-34)
http://www.microarch.org/micro34
Austin, Texas, December 1-2, 2001 Topics of interest include, but are not limited to:
Hardware/Compiler techniques for improving memoryperformance of media and stream-based processing
Application-specific hardware architectures for graphics, video,audio, communications, and other media and streamingapplications
System-on-a-chip architectures for media & stream processors Hardware/Software Co-Design of media and stream processors and others ....
Berkeley: Chip-in-a-Day Bee Project
-
7/30/2019 Tampere 01 Course 4
47/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
47
y p y j
Chip-in-a-Day Project. Prof. Dr. Bob Broderson, BWRD: targeting
a radical rethink of the ASIC design flow aimed at shorteningdesign time. Relying on stream-based DPU arrays (not rDPU andrelated EDA tools. Davis: ... 50x decrease in power requ. overtypical TI C64X design.
New design flow to break up the highly iterative EDA process,allowing designers to spend more time defining the device andfar less time implementing it in silicon. ... developers to start bycreating data flow graphs rather than C code,
It is stream-based computing by DPU array (hardwired DPA)
For hardwired and reconfigurable DPU array and rDPU array
Stanford thru BYU
-
7/30/2019 Tampere 01 Course 4
48/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
48
Stanford: Prof. Flynn went emeritus, Oskar Menzer moved to Bell Lab
no activities seen other than YAFA (yet another FPGA application) UCLA: Prof. Jason Cong, expert on FPGA architectures and R& Palgorithms. 9 projects, mult. sponsors under California MICRO Progra
Prof. Majid Sarrafzadeh directs the SPS project: "versatile IPs, a nerouting architecture, architecture-aware CAD, IP-aware SPS compiler
USC: Prof. Viktor Prasanna (EE dept.) works 20% on reconfigurablecomputing: MAARC project, DRIVE project and Efficient Self-Reconfiguration. - Prof. Dubois: RPM Project, FPGA-based emulationscalable multiprocessors.
DEFACTO proj.: compilation - architecture-independent at all levels MIT.MATRIX web pages removed `99. RAW project: a conglomerat VT. Prof. Athanas: Jbits API f. internet RTR logic ($2.7 mio DARPA).
Prof. Brad Hutchings, BYU on programming approaches for RTR Syste BYU. Prof. Brad Hutchings works on the JHDL (JAVA Hardware
Description Language) and compilation of JHDL sources into FPGAs.
Toronto thru Karlsruhe
-
7/30/2019 Tampere 01 Course 4
49/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
49
U. Toronto. Prof. J.Rose, expert in FPGA architectures and R & P alg.
The group has dev. Transmogrifier C, a C compiler creating netlist forXilinx XC4000 and Altera's Flex 8000 and Flex 10000 series FPGAs.
Founder of Right Track CAD Corporation acquired by Altera in 1999
Los Alamos National Laboratory, Los Alamos, New Mexico (JeffArnold) Project Streams-C: programming FPGAs from C sources.
Katholic University of Leuven, and IMEC: Prof. Rudy Lauwereins,methods for MPEG-4 like multimedia applications on dynamicallyreconfigurable platforms, & on reconf. instruction set processors.
University of Karlsruhe. Prof. Dr.-Ing. Juergen Becker:
hardware/software co-design, reconfigurable architectures & rel.synthesis for future mobile communication systems & synthesis w.
distributed internet-based CAD methods, partitioning co-compilers
>> ASICs dead ?
-
7/30/2019 Tampere 01 Course 4
50/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
50
Configware Market
FPGA Market
Embedded Systems (Co-Design)
Hardwired IP Cores on Board
Run-Time Reconfiguration (RTR)
Rapid Prototyping & ASIC Emulation
Evolvable Hardware (EH)
Academic Expertise
ASICs dead ?
Soft CPU
HLLs
Problems to be solved
X t L b
(When) Will FPGAs Kill ASICs?
-
7/30/2019 Tampere 01 Course 4
51/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
51
[Jonathan Rose]
ASICs Are Already Dead
They Just Dont Know It Yet!
My Position[Jonathan Rose]
X t L bWhy? [Jonathan Rose]
-
7/30/2019 Tampere 01 Course 4
52/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
52
y [ ]
1. You have to fabricate an ASIC Very hard, getting harder
2. An FPGA is pre-fabricated A standard part
immense economic advantages
Xputer Lab
Making ASICs is Damn Difficult
-
7/30/2019 Tampere 01 Course 4
53/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
53
[Jonathan Rose]
Testing Yield Cross Talk Noise Leakage Clock Tree Design Horrible very deep submicron effects we
dont even know about yet
Xputer Lab
Did I Mention Inventory? [Jonathan Rose]
-
7/30/2019 Tampere 01 Course 4
54/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
54
ASIC users must predict # parts 2 or 3 months in advance!
Never guess the Right Amount Make Too Many You Pay holding costs
Make Too Few Competitor gets the Sale
[Jonathan Rose]
Xputer Lab
[Jonathan Rose] FPGAs Give You
-
7/30/2019 Tampere 01 Course 4
55/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
55
Instant Fabrication Get to Market Fast Fix em quick
Zero NRE Charges Low Risk
Low Cost at good volume
Xputer LabFPGAs: Too Pricey & Too Slow ?
-
7/30/2019 Tampere 01 Course 4
56/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
56
[Jonathan Rose]
Custom IC Designer Can Make Logic 20x Faster,
20x Smaller than Programmable
9 Times Out of 10 You make can the thing fast by breaking it
into multiple parallel slower pieces
Xputer Lab
Whats Wrong with This Picture?
-
7/30/2019 Tampere 01 Course 4
57/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
XputerLab
57
1. Still Have to Make the Chip2. Need Two Sets of Software to Build It
The ASIC Flow
The PLD Flow
3. Have No Idea What to Connect the PLD Pins to Chances Are, You Are Going to Get It Wrong!
EmbeddedFPGA Fabric
[Jonathan Rose]
What About PLDCores on ASICs ?
XputerLab
Whats Right with This Picture!
-
7/30/2019 Tampere 01 Course 4
58/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
p
58
1. Pre-Fabricated
2.One CAD Tool Flow!
3.Can Connect Anything to Anything PLDs are built for general connectivity
Embedded
CPU Serial Link,Analog, etc.
[Jonathan Rose]
XputerLab
>> Soft CPU
-
7/30/2019 Tampere 01 Course 4
59/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
p
59
Configware Market
FPGA Market
Embedded Systems (Co-Design)
Hardwired IP Cores on Board
Run-Time Reconfiguration (RTR)
Rapid Prototyping & ASIC Emulation
Evolvable Hardware (EH)
Academic Expertise
ASICs dead
Soft CPU
HLLs Problems to be solved
XputerLab
Free 32 bit processor core
-
7/30/2019 Tampere 01 Course 4
60/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
60
XputerLab
Processors in PLDs: Excalibur
-
7/30/2019 Tampere 01 Course 4
61/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
61
High-Speed
ProcessorsIntegratedwith PLDs
ARM 922TCore
General PurposePLD
Dual-PortRAM
Single-PortRAM
[Jonathan Rose]
U i it f K i l t
XputerLab
Soft CPU: new job for compilers
-
7/30/2019 Tampere 01 Course 4
62/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
62
softCPU
FPGA
MemorycoreFPGA
CompilerHLL
University of Kaiserslautern
XputerLab
Some soft CPU core examples
-
7/30/2019 Tampere 01 Course 4
63/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
63
Spartan-II16 bit DSPDSPuva16
FLEX10K30or EPF6016
i8080AMy80
32-bitgr1050
16-bitgr1040
Altera Mercury
8 bitNios
Altera22 D-MIPS
32-bitinstr. setNios50 MHz
Altera
Mercury
16-bitinstr. set
Nios
Xilinx up to
100 on oneFPGA
32 bit
standard RISC32 reg. by 32LUT RAM-based reg.
MicroBlaze
125 MHz 70D-MIPS
platformarchitecturecore
SpartanXLRISC integer Cxr16
old Xilinx FPGABoard
16-bit RISC,2 opd. Instr.
YARD-1A
1 Flex 10K20Acorn-1
Altera, Lattice,Xilinx
8 bit CISC1Popcorn-1
Lattice4 isp30256,4 isp1016
12 bit DSPReliance-1
2 XILINX3020 LCA
8 bits Instr. +ext. ROM
REGIS
200 XC4000ECLBs
CISC, 32 reg.uP1232 8-bit
ARMARM7 clone
SPARCLeon
25 Mhz
platformarchitecturecore
University of Kaiserslautern
XputerLab
Nios Architecture (Altera)
-
7/30/2019 Tampere 01 Course 4
64/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
64
University of Kaiserslautern
XputerLab
free DSP or Processor Cores
-
7/30/2019 Tampere 01 Course 4
65/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
65
VHDL
VHDL
VHDL
VHDL
VHDL
VHDL
VHDL
VHDL
VHDL
Mentor GraphicsVHDLanother i8051 clonei8051
Synopsys8-bit micro-controlleri8051
AMD 2910 bit sliceAMD 2910
AMD 2901 4-bit sliceAMD 2901
i8085 cloneGL85
Generic 32-bit RISC CPUDLX2
SynopsysGeneric 32-bit RISC CPUDLX
6502 compatible coreFree-6502
Xilinx XC4000A 16-bit Harvard DSP with5 pipeline stages.
16-bit DSP
Max2PlusII+ 1 Altera 10k20small 8 bit CISCAcorn 1
1 Lattice CPLD isp3256-90Verilogsmall 8 bit CISCPopCorn 1
Viewlogic 7 Lattice CPLDsSchematic12bit DSP and peripheralsReliance 1
ImplementationLanguageDescriptionCPU core
University of Kaiserslautern
XputerLab
FPGA CPUs in teaching andacademic research
-
7/30/2019 Tampere 01 Course 4
66/81
2001,[email protected] http://www.fpl.uni-kl.de
University of Kaiserslautern
66
academic research
UCSC: 1990! Mraldalen University,
Eskilstuna, Sweden Chalmers University,
Gteborg, Sweden Cornell University Gray Research Georgia Tech
Hiroshima CityUniversity, Japan
Michigan State Universidad de
Valladolid, Spain Virginia Tech
WashingtonUniversity, St. Louis New Mexico Tech UC Riverside
Tokai University,Japan
University of Kaiserslautern
XputerLab
Xilinx 10Mg, 500Mt, .12 mic
-
7/30/2019 Tampere 01 Course 4
67/81
2001,[email protected] http://www.fpl.uni-kl.de
y
67
University of Kaiserslautern
XputerLab
Soft rDPA feasible ?
-
7/30/2019 Tampere 01 Course 4
68/81
2001,[email protected] http://www.fpl.uni-kl.de68
[ la S. Guccione]
University of Kaiserslautern
XputerLab
Array I/O examples
-
7/30/2019 Tampere 01 Course 4
69/81
2001,[email protected] http://www.fpl.uni-kl.de69
data streams, or, from/toembedded memory banks
datastreams,
or,from/to
embedded
memorybanks
1
10
100
1000Performance
1980 1990 2000
Proc60%/yr..
DRAM7%/yr..
Processor-MemoryPerformance Gap:(grows 50% / year)
DRAM
CPU
[ la S. Guccione]
University of Kaiserslautern
XputerLab
HLL 2 Soft Array
-
7/30/2019 Tampere 01 Course 4
70/81
2001,[email protected] http://www.fpl.uni-kl.de70
Memorysoft CPU
miscellanous
HLL Compiler
[ la S. Guccione]
University of Kaiserslautern
XputerLab
HLL 2 flex rDPA
-
7/30/2019 Tampere 01 Course 4
71/81
2001,[email protected] http://www.fpl.uni-kl.de71
MemoryCPU
miscellanous
HLL Compiler
[ la S. Guccione]
University of Kaiserslautern
XputerLab
>> HLLs
-
7/30/2019 Tampere 01 Course 4
72/81
2001,[email protected] http://www.fpl.uni-kl.de72
Configware Market
FPGA Market Embedded Systems (Co-Design)
Hardwired IP Cores on Board
Run-Time Reconfiguration (RTR)
Rapid Prototyping & ASIC Emulation
Evolvable Hardware (EH)
Academic Expertise
ASICs dead
Soft CPU
HLLs Problems to be solved
University of Kaiserslautern
XputerLab
HLLs for Hardware Design vs.System Design vs. RTR System
-
7/30/2019 Tampere 01 Course 4
73/81
2001,[email protected] http://www.fpl.uni-kl.de73
y g yDesign
HLL Compiler
System Design
CompilerHLL
RTR System Design[ la S. Guccione]
University of Kaiserslautern
XputerLab
HLLs for Hardware Design vs.System Design vs. RTR System
-
7/30/2019 Tampere 01 Course 4
74/81
2001,[email protected] http://www.fpl.uni-kl.de74
y g yDesign
HLL Compiler
System Design
CompilerHLL
RTR System Design
CompilerHLL
[ la S. Guccione]
University of Kaiserslautern
XputerLab
CPU and memory on Chip
-
7/30/2019 Tampere 01 Course 4
75/81
2001,[email protected] http://www.fpl.uni-kl.de75
CPU
core
FPGA core
Memory
coreCompilerHLL
CompilerHLL
RTR System Design
[ la S. Guccione]
University of Kaiserslautern
XputerLab
Jbit Environment
-
7/30/2019 Tampere 01 Course 4
76/81
2001,[email protected] http://www.fpl.uni-kl.de76
RTP Core
Library
JRoute
API
DeviceSimulator
UserCode
BoardScope
Debugger
XHWIF
JBits
API
TCP/IP
[ la S. Guccione]
University of Kaiserslautern
XputerLab
HLLs for Hardware Design vs.System Design vs. RTR System
-
7/30/2019 Tampere 01 Course 4
77/81
2001,[email protected] http://www.fpl.uni-kl.de77
Design
CompilerHLL
HLL Compiler
System Design
[ la S. Guccione]
University of Kaiserslautern
XputerLab
Embedded System Design
-
7/30/2019 Tampere 01 Course 4
78/81
2001,[email protected] http://www.fpl.uni-kl.de78
HLL Compiler
CPUcore
FPGA core
Memorycore
HLL Compiler
softCPU FPGA
MemorycoreFPGA
[ la S. Guccione]
University of Kaiserslautern
XputerLab
>> Problems to be solved
-
7/30/2019 Tampere 01 Course 4
79/81
2001,[email protected] http://www.fpl.uni-kl.de79
Configware Market
FPGA Market Embedded Systems (Co-Design)
Hardwired IP Cores on Board
Run-Time Reconfiguration (RTR)
Rapid Prototyping & ASIC Emulation
Evolvable Hardware (EH)
Academic Expertise
ASICs dead
Soft CPU
HLLs Problems to be solved
University of Kaiserslautern
XputerLab
Why Cant Reconfig. Software Survive?
-
7/30/2019 Tampere 01 Course 4
80/81
2001,[email protected] http://www.fpl.uni-kl.de80
Resource constraints/sizes are exposed: to programmer
in low-level representation (netlist)
Design revolves around device size
Algorithmic structure Exploited parallelism
University of Kaiserslautern
XputerLab
Schedule
-
7/30/2019 Tampere 01 Course 4
81/81
2001,[email protected] http://www.fpl.uni-kl.de81
time slot
08.30 10.00 Reconfigurable Computing (RC)10.00 10.30 coffee break
10.30 12.00 Stream-based Computing for RC
12.00 14.00 lunch break14.00 15.30 Resources for RC
15.30 16.00 coffee break
16.00 17.30 FPGAs: recent developments
17.30 end of seminar: thank you for attending