Available Toolchains and Toolchain Compatibility: Difference between revisions
No edit summary |
No edit summary |
||
| Line 9: | Line 9: | ||
==Avaliable toolchains on cluster== | ==Avaliable toolchains on cluster== | ||
Below are common compiler toolchains on the Sapelo2 cluster: | |||
{| width="100%" border="1" cellspacing="0" cellpadding="2" align="left" class="wikitable sortable" | {| width="100%" border="1" cellspacing="0" cellpadding="2" align="left" class="wikitable sortable" | ||
| Line 22: | Line 22: | ||
<!--LIST BEGIN--> | <!--LIST BEGIN--> | ||
|- | |- | ||
| | | foss || 2019b || GCC/8.3.0 || OpenMPI/3.1.4 || OpenBLAS/0.3.7, ScaLAPACK/2.0.2 || FFT/3.3.8 | ||
|- | |||
| gompi || 2019b || GCC/8.3.0 || OpenMPI/3.1.4 || NA || NA | |||
|- | |- | ||
| GCC || 6.4.0-2.28, 8.3.0, 9.2.0 || GCC || NA || NA || NA | | GCC || 6.4.0-2.28, 8.3.0, 9.2.0 || GCC || NA || NA || NA | ||
|- | |- | ||
| | | GCCcore || 6.4.0, 8.3.0, 9.2.0 || GCCcore || NA || NA || NA | ||
|- | |||
<!--LIST END--> | |||
|} | |||
Below are compiler toolchains with CUDA support on the Sapelo2 cluster: | |||
{| width="100%" border="1" cellspacing="0" cellpadding="2" align="left" class="wikitable sortable" | |||
|- | |- | ||
| | ! scope="col" width="100px"| Name | ||
! scope="col" width="100px"| Version | |||
! scope="col" width="100px"| Compiler(s) | |||
! scope="col" width="100px"| MPI library | |||
! scope="col" width="100px"| Linear algebra library | |||
! scope="col" width="100px"| FFT | |||
|- | |- | ||
<!--LIST BEGIN--> | |||
|- | |- | ||
| fosscuda || 2019b || GCC/8.3.0, CUDA/10.1.243 || OpenMPI/3.1.4 || OpenBLAS/0.3.7, ScaLAPACK/2.0.2 || FFT/3.3.8 | | fosscuda || 2019b || GCC/8.3.0, CUDA/10.1.243 || OpenMPI/3.1.4 || OpenBLAS/0.3.7, ScaLAPACK/2.0.2 || FFT/3.3.8 | ||
|- | |||
| gcccuda || 2019b || GCC/8.3.0, CUDA/10.1.243 || NA || NA || NA | |||
|- | |- | ||
<!--LIST END--> | <!--LIST END--> | ||
Revision as of 12:40, 22 June 2020
Note: This page is for using new queueing system on the Sapelo2 cluster. This page is still under development as of June 22-26, 2020.
If you are current Sapelo2 users, please come back later.
We use EasyBuild to build most of the software modules on the Sapelo2 cluster. Easybuild is a software build and installation framework that allows us to manage scientific software on High Performance Computing (HPC) systems in an efficient way. EasyBuild employs so-called compiler toolchains or, simply toolchains for short, in handling the build and installation processes. A typical toolchain consists of one or more compilers, usually put together with some libraries for specific functionality, e.g., for using an MPI stack for distributed computing, or which provide optimized routines for commonly used math operations, e.g., the well-known BLAS/LAPACK APIs for linear algebra routines. For detailed info about EasyBuild and its toolchains, please refer to EasyBuild and toolchains
Avaliable toolchains on cluster
Below are common compiler toolchains on the Sapelo2 cluster:
| Name | Version | Compiler(s) | MPI library | Linear algebra library | FFT |
|---|---|---|---|---|---|
| foss | 2019b | GCC/8.3.0 | OpenMPI/3.1.4 | OpenBLAS/0.3.7, ScaLAPACK/2.0.2 | FFT/3.3.8 |
| gompi | 2019b | GCC/8.3.0 | OpenMPI/3.1.4 | NA | NA |
| GCC | 6.4.0-2.28, 8.3.0, 9.2.0 | GCC | NA | NA | NA |
| GCCcore | 6.4.0, 8.3.0, 9.2.0 | GCCcore | NA | NA | NA |
Below are compiler toolchains with CUDA support on the Sapelo2 cluster:
| Name | Version | Compiler(s) | MPI library | Linear algebra library | FFT |
|---|---|---|---|---|---|
| fosscuda | 2019b | GCC/8.3.0, CUDA/10.1.243 | OpenMPI/3.1.4 | OpenBLAS/0.3.7, ScaLAPACK/2.0.2 | FFT/3.3.8 |
| gcccuda | 2019b | GCC/8.3.0, CUDA/10.1.243 | NA | NA | NA |
For a complete list of EasyBuild toolchains, please refer to List_of_known_toolchains