# \#mpi

**URL:** https://forum.access-hive.org.au/tag/mpi/60.md

[Latest](https://forum.access-hive.org.au/latest.md) · [Categories](https://forum.access-hive.org.au/categories.md) · [Tags](https://forum.access-hive.org.au/tags.md)

---

## [Seeking help from MPI+GPU pros!](https://forum.access-hive.org.au/t/seeking-help-from-mpi-gpu-pros/5424)

<div class="topic-metadata">

**Author:** [@taimoorsohail](https://forum.access-hive.org.au/u/taimoorsohail)\
**Replies:** 12\
**Last updated:** [7 November 2025 04:41 UTC](https://forum.access-hive.org.au/t/seeking-help-from-mpi-gpu-pros/5424 "2025-11-07T04:41:14Z")

</div>

Hi all, I am currently running a multi-GPU ocean model (ClimaOcean) and am getting very poor scaling as I increase the number of GPUs. I am investigating the performance using NVIDIA NSYS profiles, but am struggling to …

---

## ["Run ACCESS-ESM" fails with error code 139](https://forum.access-hive.org.au/t/run-access-esm-fails-with-error-code-139/1749)

<div class="topic-metadata">

**Author:** [@paulleopardi](https://forum.access-hive.org.au/u/paulleopardi)\
**Replies:** 17\
**Last updated:** [8 February 2024 00:16 UTC](https://forum.access-hive.org.au/t/run-access-esm-fails-with-error-code-139/1749 "2024-02-08T00:16:31Z")

</div>

As a step towards creating a Spack build for ACCESS-ESM1.5, I have been trying to use the “Run ACCESS-ESM1.5” instructions. I have tried to run ESM1.5 according to these instructions at least 4 times, and each time I see…

---

## [Performance of MPI in singularity/apptainer](https://forum.access-hive.org.au/t/performance-of-mpi-in-singularity-apptainer/1352)

<div class="topic-metadata">

**Author:** [@Aidan](https://forum.access-hive.org.au/u/Aidan)\
**Replies:** 0\
**Last updated:** [8 September 2023 12:31 UTC](https://forum.access-hive.org.au/t/performance-of-mpi-in-singularity-apptainer/1352 "2023-09-08T12:31:54Z")

</div>

This is an interesting article where they attempt to benchmark containerised MPI In some cases they actually seem to see improved performance for containerised MPI. Chatting with @angus-g the other day, it seems cont…

---

## [Splicing: Swapping ABI compatible libraries with spack](https://forum.access-hive.org.au/t/splicing-swapping-abi-compatible-libraries-with-spack/242)

<div class="topic-metadata">

**Author:** [@Aidan](https://forum.access-hive.org.au/u/Aidan)\
**Replies:** 0\
**Last updated:** [6 December 2022 04:57 UTC](https://forum.access-hive.org.au/t/splicing-swapping-abi-compatible-libraries-with-spack/242 "2022-12-06T04:57:41Z")

</div>

This is interesting. Using spack to swap out ABI compatible libraries, like MPI, to avoid expensive and potentially unnecessary builds and go further with a compatibility model for other aspects of ABI.

---

## [CABLE profiling results](https://forum.access-hive.org.au/t/cable-profiling-results/654)

<div class="topic-metadata">

**Author:** [@Juergen](https://forum.access-hive.org.au/u/Juergen)\
**Replies:** 1\
**Last updated:** [4 April 2023 07:34 UTC](https://forum.access-hive.org.au/t/cable-profiling-results/654 "2023-04-04T07:34:36Z")

</div>

We ran the ‘scorep’ profiler for a mpi test run of the CABLE-POP model (aka the Canberra version). The profiler checks how much time the model spends in each subroutine. It showed that more than 80% of run time was used …