General Page
Available from NVIDIA, the NVIDIA CUDA Toolkit has support for IBM® Power® Little Endian (POWER9™), and is required when you compile or link programs that use OpenMP 4.5 to offload computation to the NVIDIA GPUs.
XL C/C++ for Linux 16.1.1 Fix Pack 8 and XL Fortran for Linux 16.1.1 Fix Pack 8 fully support NVIDIA CUDA Toolkit 11.0, released in May of 2020. Install NVIDIA CUDA Toolkit 11.0 for XL C/C++ for Linux 16.1.1 Fix Pack 8 and XL Fortran for Linux 16.1.1 Fix Pack 8 to compile programs that contain code used to offload computation to the NVIDIA GPUs.
For previous compiler releases, install the following version of the NVIDIA CUDA Toolkit:
- If your compiler version is from 16.1.1 Fix Pack 3 to 16.1.1 Fix Pack 7, install NVIDIA CUDA Toolkit 10.1.
- If your compiler version is from 16.1.1 to 16.1.1 Fix Pack 2, install NVIDIA CUDA Toolkit 9.2.
To install the NVIDIA CUDA Toolkit, use the Package Manager installation. For instructions about the Package Manager installation, see NVIDIA CUDA Installation Guide for Linux.
Note: The Runfile installation is currently not supported on Power processors.
Related information:
[{"Business Unit":{"code":"BU058","label":"IBM Infrastructure w\/TPS"},"Product":{"code":"SSXVZZ","label":"XL C\/C++ for Linux"},"ARM Category":[{"code":"a8m0z00000007u2AAA","label":"C\/C++->Compiler"}],"ARM Case Number":"","Platform":[{"code":"PF016","label":"Linux"}],"Version":"16.1.1","Line of Business":{"code":"LOB57","label":"Power"}},{"Line of Business":{"code":"LOB08","label":"Cognitive Systems"},"Business Unit":{"code":"BU058","label":"IBM Infrastructure w\/TPS"},"Product":{"code":"SSAT4T","label":"XL Fortran for Linux"},"ARM Category":[{"code":"a8m0z000000088sAAA","label":"Fortran->Compiler"}],"Platform":[{"code":"PF016","label":"Linux"}],"Version":"16.1.1"}]
Was this topic helpful?
Document Information
Modified date:
01 November 2020
UID
ibm16233278