BabelStream/src
Tom Deakin 092ee67764 Change CUDA DOT thread-blocks to 1024
This improves the performance on Ampere (A100) GPUs.

Fixes #137.
2023-06-12 15:51:13 +01:00
..
acc [WIP] Drop CL headers and Makefiles 2021-11-30 18:22:55 +00:00
cuda Change CUDA DOT thread-blocks to 1024 2023-06-12 15:51:13 +01:00
fortran BabelStream Fortran 2022-11-15 14:29:56 +02:00
hip Round up dot_num_blocks and remove extra check 2023-03-13 10:47:37 -05:00
java/java-stream Version bump standalone implementations 2021-12-22 13:05:29 +00:00
julia/JuliaStream.jl Version bump standalone implementations 2021-12-22 13:05:29 +00:00
kokkos Fix compatibility with Kokkos 4+ 2023-06-10 21:08:29 +01:00
legacy [WIP] Drop CL headers and Makefiles 2021-11-30 18:22:55 +00:00
ocl Fetch CL headers on-demand for ComputeCpp and OCL build 2021-12-01 16:19:50 +00:00
omp [WIP] Drop CL headers and Makefiles 2021-11-30 18:22:55 +00:00
raja In-package linking to RAJA 2022-09-15 11:32:23 +01:00
rust/rust-stream Bump rust-stream dependencies 2022-10-01 04:56:03 +08:00
scala/scala-stream Version bump standalone implementations 2021-12-22 13:05:29 +00:00
std-data ammend copyright headers 2021-12-22 12:33:49 +00:00
std-indices Fix missing counting iterator operators for stdpar 2022-03-31 14:50:10 +01:00
std-ranges Split implementation for index/data oriented std variants 2021-12-15 21:38:56 +00:00
sycl Move SYCL2020 to subdirectory 2021-12-09 11:40:30 +00:00
sycl2020 Added ICPX support for the OneAPI2023 and later versions since DPCPP is deprecated 2023-01-27 14:28:13 +00:00
tbb Fetch CL headers on-demand for ComputeCpp and OCL build 2021-12-01 16:19:50 +00:00
thrust Merge pull request #135 from jeffhammond/fortran-ports 2022-11-23 00:01:04 +00:00
.gitignore BabelStream Fortran 2022-11-15 14:29:56 +02:00
ci-prepare-bionic.sh Sync CUDA version with CI runner 2022-04-08 05:34:15 +01:00
ci-test-compile.sh Fix Thrust/CUB path 2022-04-08 19:43:15 +01:00
main.cpp Use long double for check solution in case of very large problem sizes 2022-04-01 10:51:56 +01:00
Stream.h Use ./src instead of ./cpp 2021-05-26 17:46:07 +01:00