Answers are grounded in this paper and cite their sources.
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.
No takes yet. Share an insight, caveat, or question.
Moreau et al. (2019) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: