A Hardware–Software Blueprint for Flexible Deep Learning Specialization
Thierry Moreau, Tianqi Chen, Luis Vega, Jared Roesch, Eddie Yan, Lianmin Zheng, Josh Fromm, Ziheng Jiang, Luís Ceze, Carlos Guestrin, Arvind Krishnamurthy · IEEE Micro · 2019
This article describes the Versatile Tensor Accelerator (VTA), a programmable DL architecture designed to be extensible in the face of evolving workloads. VTA achieves “flexible specialization” via a parameterizable architecture, two-level Instruction Set Architecture (ISA), and a Just in Time (JIT) compiler.