parsec
parsec copied to clipboard
ICLDisco
PaRSEC is a generic framework for architecture aware scheduling and management of micro-tasks on distributed, GPU accelerated, many-core heterogeneous architectures. PaRSEC assigns computation threads...
Import the script developed by @evaleev and Ashawini Chudaman for TTG in https://github.com/TESSEorg/ttg/commits/master/bin/pbt_to_ctf.py and adapt it for broader use in PaRSEC
## Describe the bug PaRSEC reduces the overhead of task activations by only sending a single activation message per destination process; if a process needs multiple output data from a...
**[Original report](https://bitbucket.org/icldistcomp/parsec/issue/124) by Damien Genet (Bitbucket: [dgenet](https://bitbucket.org/dgenet), GitHub: [dgenet](https://github.com/dgenet)).** ---------------------------------------- We discussed the API for PaRSEC before the 2.0.0 release. ``` #!c /** Struct **/ /* parsec_info_t: hashmap storing */...
## Description Software depending on PaRSEC may want to check whether the correct accelerator component has been enabled. ### Describe the solution you'd like PaRSEC could provide `PaRSEC::cuda` and `PaRSEC::hip`...
PR #613 made all CI tests initialize the GPU if there is a GPU available. When running in oversubscribe mode, this can lead to falsely failing tests, that fail not...
If I simulate being unable to allocate memory on the device, both for data and for streams, I get the following stack: ``` #5 0x00007ffff7e57995 in parsec_list_destruct (list=0x7ffff7fbd2a0 ) at...
## Describe the bug Seen only once on #321, need to see if it also happens on master ### To Reproduce ``` 29302 Command: "/apps/spacks/2023-08-14/opt/spack/linux-rocky9-x86_64/gcc-11.3.1/openmpi-4.1.5-2rgaqk2wseegpmbdbbygvwrljccjaqsk/bin/mpiexec" "-n" "4" "dsl/did/dtd_test_task_insertion" ... 29329...
## Describe the bug We see the following warnings (followed by asserts in debug mode) when memory on the device is tight: ``` W@00000 GPU[hip(0)]: Write access to data copy...
## Description Some time ago we changed `--mca runtime_bind_thread` to `--mca bind_thread`. Passing in the obsolete value is silently ignored, which is really error prone for end users. Learning from...