Low Cost SIMD Module for ML Acceleration Marc Solé Bonet Leonidas Kosmidis
LowCost SIMD Module for ML Acceleration Marc Solé Bonet Leonidas Kosmidis, Universitat Politècnica de Catalunya (UPC) and Barcelona Supercomputing Center (BSC) In this presentation, Marc and Leonidas, talk about a novel approach in the acceleration of ML operations in area and powerconstrained RISCV processors. As an alternative to conventional approaches which build on the Vextension and require an additional areahungry vector register file and the implementation of several instructions for compliance, they propose an open source short SIMD module for Machine Learning acceleration which reuses the integer register file, resulting in considerable area savings. The implementation complexity is minimal since it only requires 17 new highly configurable custom instructions which combine classic SIMD operations with subsequent reduction operations over a two stage design similar to a MAC unit, which can implement 200 different combinations of commonly used ML operations. The design has been driven by the
|
|