1. What's New
Welcome to the 22.9 version of the NVIDIA HPC SDK, a comprehensive suite of compilers and libraries enabling developers to program the entire HPC platform, from the GPU foundation to the CPU and out through the interconnect. The 22.9 release of the HPC SDK is primarily a maintenance and bugfix release containing important functionality and performance improvements.
The HPC Compilers added support for PTX JIT compilation for relocatable device code mode which is the compilers’ default when generating code for NVIDIA GPUs. For more information about using this feature, please refer to the HPC Compilers documentation.
The HPC SDK now offers signed packages for download. Be sure to refer to the Developer Zone download page for further instructions on using these packages.
2. Release Component Versions
The NVIDIA HPC SDK 22.9 release contains the following versions of each component:
Linux_x86_64 | Linux_ppc64le | Linux_aarch64 | |||||||
---|---|---|---|---|---|---|---|---|---|
CUDA 10.2 | CUDA 11.0 | CUDA 11.7 | CUDA 10.2 | CUDA 11.0 | CUDA 11.7 | CUDA 10.2 | CUDA 11.0 | CUDA 11.7 | |
nvc++ | 22.9 | 22.9 | 22.9 | ||||||
nvc | 22.9 | 22.9 | 22.9 | ||||||
nvfortran | 22.9 | 22.9 | 22.9 | ||||||
nvcc | 10.2.89 | 11.0.221 | 11.7.99 | 10.2.89 | 11.0.221 | 11.7.99 | N/A | 11.0.221 | 11.7.99 |
NCCL | 2.13.4 | 2.13.4 | 2.13.4 | 2.13.4 | 2.13.4 | 2.13.4 | N/A | 2.13.4 | 2.13.4 |
NVSHMEM | 2.6.0 | 2.6.0 | 2.6.0 | 2.6.0 | 2.6.0 | 2.6.0 | N/A | N/A | N/A |
cuBLAS | 10.2.2.89 | 11.2.0.252 | 11.10.3.66 | 10.2.2.89 | 11.2.0.252 | 11.10.3.66 | N/A | 11.2.0.252 | 11.10.3.66 |
cuFFT | 10.1.2.89 | 10.2.1.245 | 10.7.2.91 | 10.1.2.89 | 10.2.1.245 | 10.7.2.91 | N/A | 10.2.1.245 | 10.7.2.91 |
cuFFTMp | N/A | N/A | 10.8.1 | N/A | N/A | 10.8.1 | N/A | N/A | N/A |
cuRAND | 10.1.2.89 | 10.2.1.245 | 10.2.10.91 | 10.1.2.89 | 10.2.1.245 | 10.2.10.91 | N/A | 10.2.1.245 | 10.2.10.91 |
cuSOLVER | 10.3.0.89 | 10.6.0.245 | 11.4.0.1 | 10.3.0.89 | 10.6.0.245 | 11.4.0.1 | N/A | 10.6.0.245 | 11.4.0.1 |
cuSOLVERMp | N/A | N/A | 0.2.1 | N/A | N/A | N/A | N/A | N/A | N/A |
cuSPARSE | 10.3.1.89 | 11.1.1.245 | 11.7.4.91 | 10.3.1.89 | 11.1.1.245 | 11.7.4.91 | N/A | 11.1.1.245 | 11.7.4.91 |
cuTENSOR | 1.6.0 | 1.6.0 | 1.6.0 | 1.6.0 | 1.6.0 | 1.6.0 | N/A | 1.6.0 | 1.6.0 |
Nsight Compute | 2022.2.0 | 2022.2.0 | 2022.2.0 | ||||||
Nsight Systems | 2022.3.4.34 | 2022.3.4.34 | 2022.3.4.34 | ||||||
OpenMPI | 3.1.5 | 3.1.5 | 3.1.5 | ||||||
HPC-X | N/A | 2.12 | 2.12 | N/A | N/A | N/A | N/A | 2.12 | 2.12 |
UCX | N/A | 1.14.0 | 1.14.0 | N/A | N/A | N/A | N/A | 1.14.0 | 1.14.0 |
OpenBLAS | 0.3.20 | 0.3.20 | 0.3.20 | ||||||
Scalapack | 2.2.0 | 2.2.0 | 2.2.0 | ||||||
Thrust | 1.9.7 | 1.9.9 | 1.15.0 | 1.9.7 | 1.9.9 | 1.15.0 | N/A | 1.9.10 | 1.15.0 |
CUB | N/A | 1.9.9 | 1.15.0 | N/A | 1.9.9 | 1.15.0 | N/A | 1.9.9 | 1.15.0 |
libcu++ | N/A | 1.0.0 | 1.8.0 | N/A | 1.0.0 | 1.8.0 | N/A | 1.0.0 | 1.8.0 |
3. Supported Platforms
3.1. Platform Requirements for the HPC SDK
Architecture | Linux Distributions | Minimum gcc/glibc Toolchain | Minimum CUDA Driver |
---|---|---|---|
x86_64 |
CentOS 7.2, 7.3, 7.4, 7.5, 7.6, 7.7, 7.8, 7.9 |
C99: 4.8 |
440.33 |
ppc64le |
RHEL 7.3, 7.4, 7.5, 7.6, 7.7, 8.0, 8.1, 8.3, 8.4 |
C99: 4.8 |
440.33 |
aarch64 |
CentOS 8.0, 8.1, 8.2, 8.3, 8.4 |
C99: 4.8 |
450.36 |
Programs generated by the HPC Compilers for x86_64 processors require a minimum of AVX instructions, which includes Sandy Bridge and newer CPUs from Intel, as well as Bulldozer and newer CPUs from AMD. POWER 8 and POWER 9 CPUs from the POWER architecture are supported. For the Arm architecture, the minimum required version is Arm v8.1.
The HPC Compilers are compatible with gcc and g++ and use the GCC C and C++ libraries; the minimum compatible versions of GCC are listed in Table 2. The minimum system requirements for CUDA and NVIDIA Math Library requiremetns are available in the NVIDIA CUDA Toolkit documentation.
3.2. Supported CUDA Toolchain Versions
The NVIDIA HPC SDK uses elements of the CUDA toolchain when building programs for execution with NVIDIA GPUs. Every HPC SDK installation package puts the required CUDA components into an installation directory called [install-prefix]/[arch]/[nvhpc-version]/cuda.
An NVIDIA CUDA GPU device driver must be installed on a system with a GPU before you can run a program compiled for the GPU on that system. The NVIDIA HPC SDK does not contain CUDA Drivers. You must download and install the appropriate CUDA Driver from NVIDIA , including the CUDA Compatibility Platform if that is required.
The nvaccelinfo tool prints the CUDA Driver version in its output. You can use it to find out which version of the CUDA Driver is installed on your system.
- CUDA 10.2
- CUDA 11.0
- CUDA 11.7
4. Known Limitations
- Prior to using HPC-X, users should take care to source the hpcx-init.sh script: $ . /[install-path]/Linux_x86_64/dev/comm_libs/hpcx/hpcx-2.11/hpcx-init.sh Then, run the hpcx_load function defined by this script: $ hpcx_load These actions will set important environment variables that are needed when running HPC-X. Also, if you see the following warning from HPC-X while running an MPI job: WARNING: Open MPI tried to bind a process but failed. This is a warning only; your job will continue, though performance may be degraded. This is a known issue, and may be worked around as follows: export OMPI_MCA_hwloc_base_binding_policy=""
- Derived type objects with zero-size derived type allocatable components that are used in sourced allocation or allocatable assignment may result in a runtime segmentation violation.
- When using -stdpar to accelerate C++ parallel algorithms, the algorithm calls cannot include virtual function calls or function calls through a function pointer, cannot use C++ exceptions, can only dereference pointers that point to the heap, and must use random access iterators (raw pointers as iterators work best).
5. Deprecations and Changes
- cudaDeviceSynchronize() in CUDA Fortran has been deprecated and will be removed in a future release.
- Starting with the 21.11 version of the NVIDIA HPC SDK, the HPC-X package is no longer shipped as part of the packages made available for the POWER architecture.
- Starting with the 21.5 version of the NVIDIA HPC SDK, the -cuda option for NVC++ and NVFORTRAN no longer automatically links the NVIDIA GPU math libraries. Please refer to the -cudalib option.
- HPC Compiler support for the Kepler architecture of NVIDIA GPUs was deprecated starting with the 21.3 version of the NVIDIA HPC SDK.
- Support for the KNL architecture of multicore CPUs in the NVIDIA HPC SDK was removed in the HPC SDK version 21.3.
Notices
Notice
ALL NVIDIA DESIGN SPECIFICATIONS, REFERENCE BOARDS, FILES, DRAWINGS, DIAGNOSTICS, LISTS, AND OTHER DOCUMENTS (TOGETHER AND SEPARATELY, "MATERIALS") ARE BEING PROVIDED "AS IS." NVIDIA MAKES NO WARRANTIES, EXPRESSED, IMPLIED, STATUTORY, OR OTHERWISE WITH RESPECT TO THE MATERIALS, AND EXPRESSLY DISCLAIMS ALL IMPLIED WARRANTIES OF NONINFRINGEMENT, MERCHANTABILITY, AND FITNESS FOR A PARTICULAR PURPOSE.
Information furnished is believed to be accurate and reliable. However, NVIDIA Corporation assumes no responsibility for the consequences of use of such information or for any infringement of patents or other rights of third parties that may result from its use. No license is granted by implication of otherwise under any patent rights of NVIDIA Corporation. Specifications mentioned in this publication are subject to change without notice. This publication supersedes and replaces all other information previously supplied. NVIDIA Corporation products are not authorized as critical components in life support devices or systems without express written approval of NVIDIA Corporation.
Trademarks
NVIDIA, the NVIDIA logo, CUDA, CUDA-X, GPUDirect, HPC SDK, NGC, NVIDIA Volta, NVIDIA DGX, NVIDIA Nsight, NVLink, NVSwitch, and Tesla are trademarks and/or registered trademarks of NVIDIA Corporation in the U.S. and other countries. Other company and product names may be trademarks of the respective companies with which they are associated.