Chip Kerchner 9b51dc7972 Fixed performance issues for VSX and P10 MMA in general_matrix_matrix_product

2021-02-17 17:49:23 +00:00

..

AltiVec

Fixed performance issues for VSX and P10 MMA in general_matrix_matrix_product

2021-02-17 17:49:23 +00:00

AVX

Updated pfrexp implementation.

2021-02-17 02:23:24 +00:00

AVX512

Updated pfrexp implementation.

2021-02-17 02:23:24 +00:00

CUDA

Specialize std::complex operators for use on GPU device.

2021-01-22 18:19:19 +00:00

Default

New accurate algorithm for pow(x,y). This version is accurate to 1.4 ulps for float, while still being 10x faster than std::pow for AVX512. A future change will introduce a specialization for double.

2021-02-17 02:50:32 +00:00

GPU

Fix for broken ROCm/HIP Support

2020-12-11 16:14:57 +00:00

HIP/hcc

moving Half headers from CUDA dir to GPU dir, removing the HIP versions

2018-06-13 12:26:18 -04:00

MSA

Remove unused packet op "palign".

2020-05-07 17:14:26 -07:00

NEON

Updated pfrexp implementation.

2021-02-17 02:23:24 +00:00

SSE

Updated pfrexp implementation.

2021-02-17 02:23:24 +00:00

SVE

Updated pfrexp implementation.

2021-02-17 02:23:24 +00:00

SYCL

[SYCL Function pointer Issue]: SYCL does not support function pointer inside the kernel, due to the portability issue of a function pointer and memory address space among host and accelerators. To fix the issue, function pointers have been replaced by function objects.

2020-11-12 01:50:28 +00:00

ZVector

Include <cstdint> in one place, remove custom typedefs

2021-01-26 14:23:05 -08:00