ZLUDA as the open-source solution bringing CUDA to non-NVIDIA hardware has been seeing a nice uptick in activity the past several months for its latest take on life. The latest feature merged to ZLUDA is the all-important kernel cache to help with performance.

ZLUDA lead developer Andrzej Janik has implemented a kernel cache in order to cache PTX code on-disk to avoid subsequent recompilations. This kernel cache leverages an SQLite database and is another step (of many) to help with bettering the performance of this CUDA implementation for non-NVIDIA hardware.

The kernel cache support was built out as part of ZLUDA-preview.93 yesterday and includes new Linux and Windows binaries for those interested. More details on ZLUDA's kernel cache for those interested via this GitHub pull .