Exploring Sparse Matrix Multiplication Kernels on the Cerebras CS-3
This work provides initial kernel designs and performance characterization for sparse linear algebra on the Cerebras CS-3, which is relevant for applications like GNNs and recommendation systems, but the results are preliminary and show limitations at high sparsity.
The paper explores sparse-dense matrix multiplication (SpMM) and sampled dense-dense matrix multiplication (SDDMM) on the Cerebras CS-3 accelerator, achieving up to 100x speedup over CPU for SpMM with 90% sparse matrices, but performance degrades below CPU for >99% sparsity.
In recent years, novel AI accelerators have emerged as promising alternatives to GPU for AI model training and inference tasks. One such accelerator, the Cerebras CS-3, achieves strong performance on large model training as well as scientific applications like molecular dynamics simulations. While dense compute workloads have been thoroughly explored for the CS-3, its potential for sparse workloads has not been fully examined. Applications requiring sparse linear algebra kernels, such as GNNs, linear solvers, and recommendation systems, could achieve good performance on a dataflow accelerator like the CS-3. In this work, we explore two key sparse linear algebra kernels, sparse-dense matrix multiplication (SpMM) and sampled dense-dense matrix multiplication (SDDMM), on the Cerebras CS-3. We propose low-level CS-3 kernel designs for these operations and optimize our designs to improve I/O performance, memory footprint, and scalability to large matrices. Our evaluation examines memory footprint and SpMM/SDDMM speedup relative to CPU. The evaluation suggests that the CS-3 can outperform CPU by 100$\times$ for SpMM with 90\% sparse matrices with performance improving as sparse matrix dimensionality increases. SDDMM on CS-3 can outperform CPU 20$\times$ for 90\% sparse matrices. We additionally find that as sparsity increases to beyond 99\%, the CS-3 suffers from performance degradation that makes it slower than CPU for SpMM.