TC-GNN
Table of Contents
paper link: https://arxiv.org/pdf/2112.02052.pdf
TL;DR: This paper propose to co-densing (SpArch also propose co-densing) the sparse matrix \(A\) in SpMM \(C=A\times B\) to reduce the fragmentation when using Tensor Cores for SpMM computations on GPU.
My Digest
The co-densing could be described as a special format rewrite rules in SparseTIR, worth trying it and add an example in the documentation.