Skip to content

2022.05.00
Compare
Choose a tag to compare
@mgates3 mgates3 released this 16 Feb 00:07
· 1353 commits to master since this release
v2022.05.00
5fb5787
  • Improved performance, including:
    LU, Cholesky, QR, mixed-precision LU and Cholesky, trsm, hemm, gemm,
    eigenvalues
  • Added LU threshold pivoting
  • Added scale, add, print
  • Added row-major MPI grid order; fixes ScaLAPACK API
  • Included HIP sources in repo, to eliminate build requirement of hipify-perl
  • Fixed OpenMP issues
  • Fixed QR with low-rank local blocks
  • Added C API in CMake
  • Rewrote testers to use less memory and reduce ScaLAPACK dependency
  • Use fast residual test for BLAS routines
Assets 3
Loading