On Bit-Serial NoCs for FPGAs

We can build lightweight bit-serial FPGA NoC routers thatcost 20 LUT, 17 FF per router and operate at 800–900 MHzspeeds. Each bit-serial router implements deflection-routing on aunidirectional torus topology requiring 1b-wide connection perport. The key ideas that enable this implementation are (1)reformulation of the dimension-ordered routing (DOR) functionusing compact 1 LUT, 1 FF streaming pattern matchers, (2)compact retiming of the datapath signals into SRL16 blocks, and(3) careful FPGA layout to efficiently pack the router logic intosmall rectangular regions 2×4 SLICEs on the chip. We anticipatethese bit-serial NoCs can be used in a variety of scenariosincluding overlay support for triggered debug, lightweight controlsignal dissemination, massively-parallel bit-serial processing.

[1]  Jan Gray.  GRVI Phalanx: A Massively Parallel RISC-V FPGA Accelerator Accelerator , 2016, 2016 IEEE 24th Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM).

[2]  Greg Stitt,et al.  Revisiting Serial Arithmetic: A Performance and Tradeoff Analysis for Parallel Applications on Modern FPGAs , 2015, 2015 IEEE 23rd Annual International Symposium on Field-Programmable Custom Computing Machines.

[3]  Onur Mutlu,et al.  A case for bufferless routing in on-chip networks , 2009, ISCA '09.

[4]  André DeHon,et al.  FPGA optimized packet-switched NoC using split and merge primitives , 2012, 2012 International Conference on Field-Programmable Technology.

[5]  Nachiket Kapre,et al.  Hoplite: Building austere overlay NoCs for FPGAs , 2015, 2015 25th International Conference on Field Programmable Logic and Applications (FPL).

[6]  Nachiket Kapre,et al.  Hoplite-DSP: Harnessing the Xilinx DSP48 multiplexers to efficiently support NoCs on FPGAs , 2016, 2016 26th International Conference on Field Programmable Logic and Applications (FPL).

[7]  James C. Hoe,et al.  CONNECT: re-examining conventional wisdom for designing nocs in the context of FPGAs , 2012, FPGA '12.

[8]  Nachiket Kapre,et al.  Deflection routing for multi-level FPGA overlay NoCs , 2016, 2016 International Conference on Field-Programmable Technology (FPT).

[9]  Vaughn Betz,et al.  Design tradeoffs for hard and soft FPGA-based Networks-on-Chip , 2012, 2012 International Conference on Field-Programmable Technology.

[10]  Steven J. E. Wilton,et al.  An adaptive virtual overlay for fast trigger insertion for FPGA debug , 2015, 2015 International Conference on Field Programmable Technology (FPT).