codeplaysoftware / portBLAS

An implementation of BLAS using the SYCL open standard.
Apache License 2.0
259 stars 50 forks source link

Add TPSV, refactor TRSV and TBSV Level 2 BLAS Operators #434

Closed pgorlani closed 1 year ago

pgorlani commented 1 year ago

This patch adds an implementation of TPSV Level 2 operator compatible with TUNING_TARGET specializations for INTEL_GPU, AMD_GPU and NVIDIA_GPU.

Moreover, this patch refactors and unifies the implementations of TRSV and TBSV with TPSV.