torch sparse linear algebra library

First open-source FP8 linear solver for consumer NVIDIA GPUs — 2-3x faster than cuBLAS FP64. pip install ssblast

A pure-Python matrix library with NumPy-level linear algebra operations. Zero dependencies, readable implementations, full matrix operation suite including decompositions, solvers, and advanced linear algebra. Perfect for learning, constrained environments, or when you need transparent matrix operations without the NumPy ecosystem.

Pioneer documentation: OpenFOAM CFD GPU offloading on Intel Arc Pro B70 Pro (Battlemage Xe2). Hardware excellent (FP64 96%, kernel latency CUDA-par), software stack not ready (GPU 66% idle, no working strong preconditioner in Ginkgo 1.10 SYCL). 14 bugs documented + direct VRAM/profiling data.