Batched routines (BLAS, LAPACK, etc.) for multi-dimensional arrays. This package provide both CPU support, for GPU support see CuBatchedRoutines.

pkg> add BatchedRoutines

• batched_scal
• batched_dot
• batched_nrm2
• batched_asum
• batched_axpy!
• batched_axpby
• batched_iamax

• batched_gemv
• batched_gbmv
• batched_symv
• batched_hemv
• batched_sbmv
• batched_hbmv
• batched_trmv
• batched_trsv
• batched_ger
• batched_syr
• batched_her

• batched_gemm (TODO: use gemm_batch when mkl is available)
• batched_symm
• batched_hemm
• batched_syrk
• batched_herk
• batched_syr2k
• batched_trmm
• batched_trsm
• batched_trmm

• batched_tr

For routines (e.g gemm), we use a prefix batched_ for their corresponding routines in BLAS or LAPACK and they should only define with AbstractArray{T, 3} (batched matrix) or AbstractArray{T, 2} (batched vector).

MIT