fasttrack leveraging heterogeneous fpga wires to design
play

FastTrack : Leveraging Heterogeneous FPGA Wires to Design Low-cost - PowerPoint PPT Presentation

Jul 12, 2023 •206 likes •703 views

FastTrack : Leveraging Heterogeneous FPGA Wires to Design Low-cost High-performance Soft NoCs Nachiket Kapre + Tushar Krishna nachiket@uwaterloo.ca, tushar@ece.gatech.edu 1/29 Claim FPGA overlay NoCs designed to exploit interconnect

  1. FastTrack : Leveraging Heterogeneous FPGA Wires to Design Low-cost High-performance Soft NoCs Nachiket Kapre + Tushar Krishna nachiket@uwaterloo.ca, tushar@ece.gatech.edu 1/29

  2. Claim FPGA overlay NoCs designed to exploit interconnect properties of the FPGA fabric can surpass existing state-of-the-art NoCs by: ◮ 2.5–2.8 × throughput ↑ ◮ 2.2 × energy ↓ ◮ at 2.5 × LUT cost ↑ Xilinx Virtex-7 485T FPGA, 8 × 8 system size, synthetic+real-world traffic. 2/29

  3. Context ◮ FPGAs finding comfortable home in datacenters ◮ Offloading compute intensive workloads to the FPGA ◮ Energy-efficiency, fast coupling to networking ◮ Common Infrastructure : NoCs for apps + system IO 3/29

  4. Context ◮ FPGAs finding comfortable home in datacenters ◮ Offloading compute intensive workloads to the FPGA ◮ Energy-efficiency, fast coupling to networking ◮ Common Infrastructure : NoCs for apps + system IO 3/29

  5. Landscape of contemporary FPGA NoC Routers 4/29

  6. Landscape of contemporary FPGA NoC Routers ● Split−Merge BLESS CONNECT OpenSMART Peak S/W Bandwidth (packets/ns) 1.00 ● 0.75 0.50 0.25 0.00 0 1000 2000 3000 Cost per Switch max(LUTs,FFs) ◮ ASIC clones transplanted onto FPGAs fare poorly! → expensive buffers, virtual channels, multi-ported switches ◮ Even contemporary FPGA routers are expensive and slow ◮ FastTrack : Deflection-routing + Bufferless + Torus 5/29

  7. Landscape of contemporary FPGA NoC Routers ● Split−Merge BLESS CONNECT OpenSMART Peak S/W Bandwidth (packets/ns) 1.00 ● 0.75 0.50 0.25 0.00 0 1000 2000 3000 Cost per Switch max(LUTs,FFs) ◮ ASIC clones transplanted onto FPGAs fare poorly! → expensive buffers, virtual channels, multi-ported switches ◮ Even contemporary FPGA routers are expensive and slow ◮ FastTrack : Deflection-routing + Bufferless + Torus 5/29

  8. Landscape of contemporary FPGA NoC Routers ● Split−Merge BLESS CONNECT OpenSMART Peak S/W Bandwidth (packets/ns) 3 2 1 ● 0 0 1000 2000 3000 Cost per Switch max(LUTs,FFs) ◮ ASIC clones transplanted onto FPGAs fare poorly! → expensive buffers, virtual channels, multi-ported switches ◮ Even contemporary FPGA routers are expensive and slow ◮ FastTrack : Deflection-routing + Bufferless + Torus 5/29

  9. Landscape of contemporary FPGA NoC Routers ● Split−Merge BLESS Hoplite CONNECT OpenSMART FastTrack Peak S/W Bandwidth (packets/ns) 3 2 1 ● 0 0 1000 2000 3000 Cost per Switch max(LUTs,FFs) ◮ ASIC clones transplanted onto FPGAs fare poorly! → expensive buffers, virtual channels, multi-ported switches ◮ Even contemporary FPGA routers are expensive and slow ◮ FastTrack : Deflection-routing + Bufferless + Torus 5/29

  10. Qualitative Comparison of FPGA NoC Routers Router Cost Xbar+Arb Buffers VCs OpenSMART ✗ ✗ ✗ BLESS ✗ ✓ ✓ CONNECT ✗ ✗ ✗ Split-Merge ✗ ✗ ✓ Hoplite ✓✓ ✓ ✓ 6/29

  11. Quick Tutorial on Hoplite sw sw sw sw 0,0 1,0 2,0 3,0 sw sw sw sw 0,1 1,1 2,1 3,1 sw sw sw sw 0,2 1,2 2,2 3,2 sw sw sw sw 0,3 1,3 2,3 3,3 Hoplite: A Deflection-Routed Directional Torus NoC for FPGAs , TRETS 2017 Hoplite: Building Austere Overlay NoCs for FPGAs , FPL 2015 7/29

  12. Quick Tutorial on HopliteRT sw sw sw sw 0,0 1,0 2,0 3,0 sw sw sw sw 0,1 1,1 2,1 3,1 sw sw sw sw 0,2 1,2 2,2 3,2 sw sw sw sw 0,3 1,3 2,3 3,3 HopliteRT: An Efficient FPGA NoC for Real-Time Applications , FPT 2017 8/29

  13. Qualitative Comparison of FPGA NoC Routers Router Cost Xbar+Arb Buffers VCs OpenSMART ✗ ✗ ✗ BLESS ✗ ✓ ✓ CONNECT ✗ ✗ ✗ Split-Merge ✗ ✗ ✓ Hoplite ✓✓ ✓ ✓ 9/29

  14. Qualitative Comparison of FPGA NoC Routers Router Cost Perf Xbar+Arb Buffers VCs Tput Latency OpenSMART ✗ ✗ ✗ ✓ ✓ BLESS ✗ ✓ ✓ ✓ ✗ CONNECT ✗ ✗ ✗ ✓ ✓ Split-Merge ✗ ✗ ✓ ✓ ✓ Hoplite ✓✓ ✓ ✓ ✗ ✗ 9/29

  15. Challenge ◮ Deflection routing → inefficient use of wiring resources ◮ Deflected packets stay in network for longer → latency ↑ ◮ Steal bandwidth from other traffic → throughput ↓ ◮ Can we allow improve NoC performance under deflection routing? ◮ Are there unique opportunities provided by the FPGA fabric? ◮ Hoplite cheap in LUT cost. . . ◮ FastTrack → inspect FPGA interconnect 10/29

  16. Outline Introduction and Motivation FastTrack NoC Organization FastTrack Router Operation Evaluation 11/29

  17. Outline Introduction and Motivation FastTrack NoC Organization FastTrack Router Operation Evaluation 12/29

  18. FPGA Wire Speeds distances not to scale 13/29

  19. FastTrack NoC Organization sw sw sw sw sw sw sw sw sw sw sw sw sw sw sw sw 14/29

  20. Depopulated Topology Generation sw sw sw sw sw sw sw sw sw sw sw sw sw sw sw sw sw sw sw sw 15/29

  21. Parametric Topology generation ◮ FPGA NoC parameterized by three terms: ◮ N System size ◮ D Distance of express link ◮ R Depopulation parameter → controls how many routers are FastTrack vs. vanilla Hoplite ◮ Fully populated 4 × 4 NoC → FT(16,2,1) ◮ Half population 4 × 4 NoC → FT(16,2,2) 16/29

  22. Outline Introduction and Motivation FastTrack NoC Organization FastTrack Router Operation Evaluation 17/29

  23. FastTrack Switch Organization N Sh N Ex W Ex 5:1 E Ex N W Sh W 4:1 3:1 E Sh E 4:1 4:1 3:1 S Sh S Ex PE PE S (a) Base HopliteRT (b) FastTrack 18/29

  24. Switch Operation ◮ Packets can start in either short or express links ◮ DOR routing function: N Sh N Ex travel in X first, then Y ◮ Packets can upgrade to W Ex 5:1 E Ex fast links if they can ◮ Packets can downgrade to W Sh 4:1 E Sh slow links only on turn! ◮ Livelock avoidance: 4:1 4:1 W → S > N → S S Sh S Ex PE ◮ Express links=higher priority, deflected packets acquire higher priority → progress 19/29

  25. Switch Operation ◮ Packets can start in either short or express links ◮ DOR routing function: N Sh N Ex travel in X first, then Y ◮ Packets can upgrade to W Ex 5:1 E Ex fast links if they can ◮ Packets can downgrade to W Sh 4:1 E Sh slow links only on turn! ◮ Livelock avoidance: 4:1 4:1 W → S > N → S S Sh S Ex PE ◮ Express links=higher priority, deflected packets acquire higher priority → progress 19/29

  26. Switch Operation ◮ Packets can start in either short or express links ◮ DOR routing function: N Sh N Ex travel in X first, then Y ◮ Packets can upgrade to W Ex 5:1 E Ex fast links if they can ◮ Packets can downgrade to W Sh 4:1 E Sh slow links only on turn! ◮ Livelock avoidance: 4:1 4:1 W → S > N → S S Sh S Ex PE ◮ Express links=higher priority, deflected packets acquire higher priority → progress 19/29

  27. Switch Operation ◮ Packets can start in either short or express links ◮ DOR routing function: N Sh N Ex travel in X first, then Y ◮ Packets can upgrade to W Ex 5:1 E Ex fast links if they can ◮ Packets can downgrade to W Sh 4:1 E Sh slow links only on turn! ◮ Livelock avoidance: 4:1 4:1 W → S > N → S S Sh S Ex PE ◮ Express links=higher priority, deflected packets acquire higher priority → progress 19/29

  28. Switch Operation ◮ Packets can start in either short or express links ◮ DOR routing function: N Sh N Ex travel in X first, then Y ◮ Packets can upgrade to W Ex 5:1 E Ex fast links if they can ◮ Packets can downgrade to W Sh 4:1 E Sh slow links only on turn! ◮ Livelock avoidance: 4:1 4:1 W → S > N → S S Sh S Ex PE ◮ Express links=higher priority, deflected packets acquire higher priority → progress 19/29

  29. Switch Operation ◮ Packets can start in either short or express links ◮ DOR routing function: N Sh N Ex travel in X first, then Y ◮ Packets can upgrade to W Ex 5:1 E Ex fast links if they can ◮ Packets can downgrade to W Sh 4:1 E Sh slow links only on turn! ◮ Livelock avoidance: 4:1 4:1 W → S > N → S S Sh S Ex PE ◮ Express links=higher priority, deflected packets acquire higher priority → progress 19/29

  30. Switch Operation ◮ Packets can start in either short or express links ◮ DOR routing function: N Sh N Ex travel in X first, then Y ◮ Packets can upgrade to W Ex 5:1 E Ex fast links if they can ◮ Packets can downgrade to W Sh 4:1 E Sh slow links only on turn! ◮ Livelock avoidance: 4:1 4:1 W → S > N → S S Sh S Ex PE ◮ Express links=higher priority, deflected packets acquire higher priority → progress 19/29

  31. Switch Operation ◮ Packets can start in either short or express links ◮ DOR routing function: N Sh N Ex travel in X first, then Y ◮ Packets can upgrade to W Ex 5:1 E Ex fast links if they can ◮ Packets can downgrade to W Sh 4:1 E Sh slow links only on turn! ◮ Livelock avoidance: 4:1 4:1 W → S > N → S S Sh S Ex PE ◮ Express links=higher priority, deflected packets acquire higher priority → progress 19/29

  32. Outline Introduction and Motivation FastTrack NoC Organization FastTrack Router Operation Evaluation 20/29

Recommend

FastTrack from FastTrack SharePoint Presentation www.officelabs.co.uk By OfficeLabs CONTENTS

FastTrack from FastTrack SharePoint Presentation www.officelabs.co.uk By OfficeLabs CONTENTS

FastTrack from FastTrack SharePoint Presentation www.officelabs.co.uk By OfficeLabs CONTENTS FastTrack SharePoint What is it? Key Benefits Key Features Support Next Step FastTrack SharePoint Presentation

389 views • 7 slides
Parallel Programming and Heterogeneous Computing FPGA Accelerators Max Plauth, Sven Khler, Felix

Parallel Programming and Heterogeneous Computing FPGA Accelerators Max Plauth, Sven Khler, Felix

Parallel Programming and Heterogeneous Computing FPGA Accelerators Max Plauth, Sven Khler, Felix Eberhardt, Lukas Wenzel and Andreas Polze Operating Systems and Middleware Group FPGA Hardware Characteristics Application Specific Integrated

591 views • 10 slides
Wires Within Wires A Minimal Model for Computational Bioelectronic Peptide Design R. A. Mansbach 1

Wires Within Wires A Minimal Model for Computational Bioelectronic Peptide Design R. A. Mansbach 1

Wires Within Wires A Minimal Model for Computational Bioelectronic Peptide Design R. A. Mansbach 1 A. L. Ferguson 2 1 Physics Department 2 Materials Science Department University of Illinois at Urbana-Champaign Blue Waters Symposium, Sunriver,

603 views • 35 slides
FPGA What is a FPGA? How FPGAs work How do they work? Manufacturers

FPGA What is a FPGA? How FPGAs work How do they work? Manufacturers

2/21/2012 Content FPGA What is a FPGA? How FPGAs work How do they work? Manufacturers Distributed RAM History FPGA vs ASIC FPGA and Microprocessors Alternatives to FPGAs ETI135, Advanced Digital IC Design

123 views • 10 slides
H ETERO R EFACTOR : Refactoring for Heterogeneous Computing with FPGA Jason Lau*, Aishwarya

H ETERO R EFACTOR : Refactoring for Heterogeneous Computing with FPGA Jason Lau*, Aishwarya

H ETERO R EFACTOR : Refactoring for Heterogeneous Computing with FPGA Jason Lau*, Aishwarya Sivaraman*, Qian Zhang*, Muhammad Ali Gulzar, Jason Cong, Miryung Kim University of California, Los Angeles *Equal co-first authors in alphabetical

649 views • 42 slides
FPGA Architecture Support for Heterogeneous, Relocatable Partial Bitstreams Christophe H URIAUX v

FPGA Architecture Support for Heterogeneous, Relocatable Partial Bitstreams Christophe H URIAUX v

24th International Conferenceon Field Programmable Logic and Applications September 3 rd , 2014 FPGA Architecture Support for Heterogeneous, Relocatable Partial Bitstreams Christophe H URIAUX v , Olivier S ENTIEYS v , Russell T ESSIER

320 views • 30 slides
F BLAS: Streaming Linear Algebra Kernels on FPGA 5 TH International Workshop on Heterogeneous

F BLAS: Streaming Linear Algebra Kernels on FPGA 5 TH International Workshop on Heterogeneous

spcl.inf.ethz.ch @spcl_eth T IZIANO D E M ATTEIS , J OHANNES DE F INE L ICHT AND T ORSTEN H OEFLER F BLAS: Streaming Linear Algebra Kernels on FPGA 5 TH International Workshop on Heterogeneous High-performance Reconfigurable Computing

242 views • 10 slides
Andrew Clinton, Matt Liberty, Ian Kuon FPGA Routing (Interconnect) FPGA routing consists of a

Andrew Clinton, Matt Liberty, Ian Kuon FPGA Routing (Interconnect) FPGA routing consists of a

Andrew Clinton, Matt Liberty, Ian Kuon FPGA Routing (Interconnect) FPGA routing consists of a network of wires and programmable switches Wire is modeled with a reduced RC network Drivers are modeled as a SPICE netlist 2-Level pass

673 views • 20 slides
Parallel Programming and Heterogeneous Computing FPGA Accelerators Max Plauth, Sven Khler, Felix

Parallel Programming and Heterogeneous Computing FPGA Accelerators Max Plauth, Sven Khler, Felix

Parallel Programming and Heterogeneous Computing FPGA Accelerators Max Plauth, Sven Khler, Felix Eberhardt, Lukas Wenzel and Andreas Polze Operating Systems and Middleware Group Introduction Mapping Workloads to Hardware Example: Given Arrays

919 views • 52 slides
Balancing Graph Processing Workloads Using Work Stealing on Heterogeneous CPU-FPGA Systems Matthew

Balancing Graph Processing Workloads Using Work Stealing on Heterogeneous CPU-FPGA Systems Matthew

Balancing Graph Processing Workloads Using Work Stealing on Heterogeneous CPU-FPGA Systems Matthew Agostini , Francis OBrien and Tarek S. Abdelrahman The Edward S. Rogers Sr. Department of Electrical and Computer Engineering University of

659 views • 34 slides
Design and Verilog Module Creating a design and implementing in FPGA Design Verilog

Design and Verilog Module Creating a design and implementing in FPGA Design Verilog

Design and Verilog Module Creating a design and implementing in FPGA Design Verilog (mainly) Schematic(sometimes) Verify Need writing a testbench Simulation using ISIM Creating a design and implementing in FPGA Add User

276 views • 15 slides
FPGA language ( HDL ). Overview 3 4 Component-Based Software Design LMES Component-Based

FPGA language ( HDL ). Overview 3 4 Component-Based Software Design LMES Component-Based

11/05/2016 Agenda The topics that will be addressed are: Overview on basic characteristics of the FPGA; Scheduling tasks on Reconfigurable FPGA reconfiguration capabilities; FPGA architectures Timing analysis for reconfigurable

299 views • 10 slides
Implementing Logic in FPGA Embedded Memory Arrays: Heterogeneous Memory Architectures Steve

Implementing Logic in FPGA Embedded Memory Arrays: Heterogeneous Memory Architectures Steve

Implementing Logic in FPGA Embedded Memory Arrays: Heterogeneous Memory Architectures Steve Wilton University of British Columbia Vancouver, B.C., Canada stevew@ece.ubc.ca As FPGAs Get Bigger... Embedded Memory is becoming critical

539 views • 31 slides
FPGA security Nele Mentens nele.mentens@kuleuven.be Design and security of cryptographic

FPGA security Nele Mentens nele.mentens@kuleuven.be Design and security of cryptographic

FPGA security Nele Mentens nele.mentens@kuleuven.be Design and security of cryptographic algorithms and devices for real-world applications June 1-6, 2014, ibenik , Croatia Outline Introduction FPGA vs. ASIC FPGA application

443 views • 17 slides
T42 Transputer Design in FPGA Transputer Design in FPGA T42 Year- -Two Design Status

T42 Transputer Design in FPGA Transputer Design in FPGA T42 Year- -Two Design Status

T42 Transputer Design in FPGA Transputer Design in FPGA T42 Year- -Two Design Status Report Two Design Status Report Year a and Martin ZABEL b , Uwe MIELKE a Uwe MIELKE and Martin ZABEL b , in collaboration w/ Michael BRUESTLE c c

702 views • 22 slides
T42 Transputer Design in FPGA Transputer Design in FPGA T42 Year- -Three Design Status

T42 Transputer Design in FPGA Transputer Design in FPGA T42 Year- -Three Design Status

T42 Transputer Design in FPGA Transputer Design in FPGA T42 Year- -Three Design Status Report Three Design Status Report Year a and Martin ZABEL b , Uwe MIELKE a and Martin ZABEL b , Uwe MIELKE in collaboration w/ Michael BRUESTLE c

226 views • 19 slides
Levels in Processor Design Circuit design Keywords: transistors, wires etc.Results in

Levels in Processor Design Circuit design Keywords: transistors, wires etc.Results in

Levels in Processor Design Circuit design Keywords: transistors, wires etc.Results in gates, flip-flops etc. Logical design Putting gates (AND, NAND, ) and flip-flops together to build basic blocks such as registers, ALUs

124 views • 9 slides
The Era of SoC FPGAs Nizar Abdallah, Ph.D. Workshop on FPGA Design for Scientific Instrumentation

The Era of SoC FPGAs Nizar Abdallah, Ph.D. Workshop on FPGA Design for Scientific Instrumentation

Power Matters The Era of SoC FPGAs Nizar Abdallah, Ph.D. Workshop on FPGA Design for Scientific Instrumentation and Computing International Centre for Theoretical Physics November 2017 Outline Introduction SoC FPGA Architectures: an

938 views • 78 slides
CDA 4253 FPGA System Design FPGA Architectures Hao Zheng Dept of Comp Sci & Eng U of South

CDA 4253 FPGA System Design FPGA Architectures Hao Zheng Dept of Comp Sci & Eng U of South

CDA 4253 FPGA System Design FPGA Architectures Hao Zheng Dept of Comp Sci & Eng U of South Florida 1 How to HW Reconfigurable Not SW Change structure Change connections among components Change logic functions of components

711 views • 70 slides
Power-Driven DNN Dataflow Optimization on FPGA Qi Sun 1 , Tinghuan Chen 1 , Jin Miao 2 , Bei Yu 1 1

Power-Driven DNN Dataflow Optimization on FPGA Qi Sun 1 , Tinghuan Chen 1 , Jin Miao 2 , Bei Yu 1 1

Power-Driven DNN Dataflow Optimization on FPGA Qi Sun 1 , Tinghuan Chen 1 , Jin Miao 2 , Bei Yu 1 1 The Chinese University of Hong Kong 2 Cadence Design Systems, Inc. 1 / 22 FPGA Targeted DNN Accelerator Design Flow Dataflow Synthesis HLS Model

452 views • 22 slides
CDA 4253 FPGA System Design Op7miza7on Techniques Hao Zheng Comp S ci & Eng Univ of South

CDA 4253 FPGA System Design Op7miza7on Techniques Hao Zheng Comp S ci & Eng Univ of South

CDA 4253 FPGA System Design Op7miza7on Techniques Hao Zheng Comp S ci & Eng Univ of South Florida 1 Extracted from Advanced FPGA Design by Steve Kilts 2 Op7miza7on for Performance 3 Performance Defini7ons Throughput : the number

520 views • 41 slides
Tsunami simulation on FPGA/GPU Tsunami simulation on FPGA/GPU and its analysis based on Statistical

Tsunami simulation on FPGA/GPU Tsunami simulation on FPGA/GPU and its analysis based on Statistical

Tsunami simulation on FPGA/GPU Tsunami simulation on FPGA/GPU and its analysis based on Statistical Model Checking Masahiro Fujita VLSI Design and Education Center (VDEC) University of Tokyo 1 2 Outline What Tsunami simulation means in

1.25k views • 44 slides
ECE532 Design Project Photoshop Functionalities on FPGA Pearl Liu George Ng Initial Goals

ECE532 Design Project Photoshop Functionalities on FPGA Pearl Liu George Ng Initial Goals

ECE532 Design Project Photoshop Functionalities on FPGA Pearl Liu George Ng Initial Goals Implementation Photoshop filters in real-time on FPGA Real time image capture, filtering, and display Xilinx Multimedia Board Final Design VGA VGA

716 views • 8 slides
Public FPGA based DM Public FPGA based DMA Atta A Attacking king UlfFrisk Agenda Background

Public FPGA based DM Public FPGA based DMA Atta A Attacking king UlfFrisk Agenda Background

Public FPGA based DM Public FPGA based DMA Atta A Attacking king UlfFrisk Agenda Background and Previous work Transmit and Receive PCIe TLPs DUMP memory FPGA Design Attack vulnerable vanilla Linux system Attack vulnerable UEFI Windows

359 views • 19 slides

More recommend