High-performance CUDA kernels for CSR sparse x dense matmul with full benchmarks and PyTorch support.
-
Updated
Dec 9, 2025 - Cuda
#
sparse-matrices
Here are 2 public repositories matching this topic...
High-performance GPU-accelerated linear algebra library for scientific computing. Custom kernels outperform cuBLAS+cuSPARSE by 2.4x in iterative solvers. Built for circuit simulation workloads.
benchmarking hpc optimization linear-algebra parallel-computing cuda cublas scientific-computing matrix-multiplication gpu-acceleration cuda-kernels gpu-computing numerical-methods conjugate-gradient circuit-simulation sparse-matrices performance-optimization cusparse custom-kernels
-
Updated
Dec 6, 2025 - Cuda
Improve this page
Add a description, image, and links to the sparse-matrices topic page so that developers can more easily learn about it.
Add this topic to your repo
To associate your repository with the sparse-matrices topic, visit your repo's landing page and select "manage topics."