Optimizing Bit-Serial Matrix Multiplication for Reconfigurable Computing
Open Access
- 20 August 2019
- journal article
- research article
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Reconfigurable Technology and Systems
- Vol. 12 (3) , 1-24
Abstract
Matrix-matrix multiplication is a key computational kernel for numerous applications in science and engineering, with ample parallelism and data locality that lends itself well to high-performance implementations. Many matrix multiplication-dependent applications can use reduced-precision integer or fixed-point representations to increase their performance and energy efficiency while still offering adequate quality of results. However, precision requirements may vary between different application phases or depend on input data, rendering constant-precision solutions ineffective. BISMO, a vectorized bit-serial matrix multiplication overlay for reconfigurable computing, previously utilized the excellent binary-operation performance of FPGAs to offer a matrix multiplication performance that scales with required precision and parallelism. We show how BISMO can be scaled up on Xilinx FPGAs using an arithmetic architecture that better utilizes six-input LUTs. The improved BISMO achieves a peak performance of 15.4 binary TOPS on the Ultra96 board with a Xilinx UltraScale+ MPSoC.Keywords
All Related Versions
Funding Information
- Vetenskapsrådet (2015-05159)
This publication has 10 references indexed in Scilit:
- A Customizable Matrix Multiplication Framework for the Intel HARPv2 Xeon+FPGA Platform2018
- Advanced Compressor Tree Synthesis for FPGAsIEEE Transactions on Computers, 2018
- Generic and universal parallel matrix summation with a flexible compression goal for Xilinx FPGAs2017
- Weighted-Entropy-Based Quantization for Deep Neural Networks2017
- FINN2017
- A Survey of Techniques for Approximate ComputingACM Computing Surveys, 2016
- Pipelined compressor tree optimization using integer linear programming2014
- Chisel2012
- Compressor tree synthesis on commercial high-performance FPGAsACM Transactions on Reconfigurable Technology and Systems, 2011
- Why systolic architectures?Computer, 1982