From e1a73093c93d9e6139aeef863750c491424d7a26 Mon Sep 17 00:00:00 2001 From: OutisLi Date: Thu, 3 Apr 2025 08:58:22 +0800 Subject: [PATCH] 20250403-fix a typo(missing newline) in cuda.md --- docs/advanced/acceleration/cuda.md | 1 + 1 file changed, 1 insertion(+) diff --git a/docs/advanced/acceleration/cuda.md b/docs/advanced/acceleration/cuda.md index 32319617669..7293614746f 100644 --- a/docs/advanced/acceleration/cuda.md +++ b/docs/advanced/acceleration/cuda.md @@ -46,6 +46,7 @@ We provides [examples](https://github.com/deepmodeling/abacus-develop/tree/devel PW basis: - Only k point parallelization is supported, so the input keyword `kpar` will be set to match the number of MPI tasks automatically. - By default, CUDA architectures 60, 70, 75, 80, 86, and 89 are compiled (if supported). It can be overriden using the CMake variable [`CMAKE_CUDA_ARCHITECTURES`](https://cmake.org/cmake/help/latest/variable/CMAKE_CUDA_ARCHITECTURES.html) or the environmental variable [`CUDAARCHS`](https://cmake.org/cmake/help/latest/envvar/CUDAARCHS.html). + LCAO basis: - Unless there is a specific reason, avoid using multiple GPUs, as it can be slower than using a single GPU. This is because the generalized eigenvalue solution of the LCAO basis set will incur additional communication overhead when calculated on multiple cards. When the memory limit of a GPU card makes it insufficient to complete the task, it is recommended to use multiple cards for calculation. - When using elpa on GPUs, some ELPA internal logs will be output. \ No newline at end of file