Democratize CUDA optimizations @NVIDIA
CUDA Core Compute Libraries (CCCL)
Former lead of the Sparse Linear Algebra team | Opinions are my own.