1 / 18

Balancing Interconnect and Computation in a Reconfigurable Array

Balancing Interconnect and Computation in a Reconfigurable Array. Why you don’t really want 100% LUT utilization. Dr. André DeHon BRASS Project University of California at Berkeley. Question. How much interconnect do I need for my computing/programmable array?

mdelaney
Download Presentation

Balancing Interconnect and Computation in a Reconfigurable Array

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Balancing Interconnect and Computation in a Reconfigurable Array Why you don’t really want 100% LUT utilization Dr. André DeHon BRASS Project University of California at Berkeley

  2. Question How much interconnect do I need for my computing/programmable array? Problem(?): too little interconnect won’t be able to use all the gates/LUTs Typical subgoal: how much interconnect to use (almost) all LUTs?

  3. Wrong Subgoal • Observation: • interconnect is dominant area on FPGAs • more important to use interconnect efficiently than to use LUTs efficiently • Different question/subgoal: • What level of interconnect gives the least implementation area for applications?

  4. LUT Utilization predict Area?

  5. Outline • Question: how much interconnect? • Teaser: less than 100% LUT utilization • Model • Application characteristics • Compose • Conclusions

  6. N/2 cutsize N/2 Model Interconnect Requirements and Richness • Recursively partition (bisect) design • Look at I/O from each partition (subtree)

  7. Regularizing Growth • How do bisection bandwidths shrink (grow) at different levels of bisection hierarchy? • Basic assumption: Geometric • 1 • 1/ • 1/2

  8. Rent’s Rule • Long standing empirical relationship • IO = CNP • 0P 1.0 • Embodies geometric assumption (C,P) • Two parameters • C base of growth • P capture growth (a= 2P) • Captures notion of locality

  9. (2 1)* => a=2 (2 2 2 1)* =>a=2(3/4) (2 2 1)* => a=(2*2)(1/3) =2(2/3) Define Network with Parameters

  10. “Cartoon” VLSI Area Model (Example artificially small for clarity)

  11. 1024 LUT Area Comparison 0.25 P=0.5 0.37 P=0.67 1.00 P=0.75 Effects of P on Area

  12. Application Requirements: Benchmark Wide (MCNC)

  13. Benchmark Parameters Interconnect requirements vary across applications.

  14. Network Fixed Schedule • Network will have a fixed wiring schedule • Applications have varying requirements • To assess impact of mismatch • map to network schedules • look at area required

  15. Mapping Problem • When design interconnect exceeds network • have to repartition to meet fixed wire schedules of target network • depopulating LUTs as necessary • See paper/poster for one approach

  16. Resources  Area Model => Area

  17. Picking Network Design Point (8l wire pitch; 2500l2 switchpoints; linear population)

  18. Summary • Interconnect area dominates logic block area • Interconnect requirements vary • among designs • within a single design • To minimize area • focus on using dominant resource (interconnect) • may underuse non-dominant resources (LUTs)

More Related