Home

Jump to bottom

Bryce Allen edited this page Nov 11, 2020 · 5 revisions

Welcome to the gpublas wiki!

Notes on why this exists.

MAGMA - CUDA and HIP implementations only
SLATE - CUDA only?
heFFTe - we don't need cross-node distributed, no SYCL support