ABSTRACT

Many of today's deep neural network accelerators, e.g., Google's TPU and NVIDIA's tensor core, are built around accelerating the general matrix multiplication (i.e., GEMM). However, supporting convolution on GEMM-based accelerators is not trivial. The naive method explicitly lowers the convolution to GEMM, commonly known as im2co1, which introduces significant performance and memory overhead. Existing implicit im2co1 algorithms require unscalable hardware and are inefficient in supporting important convolution variants such as strided convolution. In this paper, we propose a memory-efficient and hardware-friendly implicit im2co1 algorithm used by Google's TPU, which dynamically converts a convolution into a GEMM with practically zero performance and memory overhead, fully unleashing the power of GEMM engines. Through comprehensive experimental results, we quantitatively argue that this algorithm has been adopted in commercial closed-source platforms, and we are the first to describe its high-level idea and implementation details. Finally, we show that our algorithm can also be generally applied to Nvidia's Tensor Cores (TC), matching and out-performing the measured performance on TCs.