-
Base material at OpenMPTargetOffload.ppt
- Data transfers overview
- map directives ( `to`, `from`,`tofrom`,` alloc`, `always` )
- `#pragma omp enter/exit/update data`
- zero copy ( unified memory ,…
-
**Description of the problem**
I'm debugging why ilastik's builds on Windows fail with pytorch 2.4 and above (https://github.com/ilastik/ilastik/pull/2927), and have been able to trace the problem ba…
-
Attempts to build GDASApp `develop` at [5fb52d3](https://github.com/NOAA-EMC/GDASApp/commit/5fb52d368aeda52b205e8323b4d2db7ff3082e70) on Cactus fail with an oops configuration error
```
-- ---------…
-
Show how to allocate mpi shared memory and pinned memory .
- [x] shared memory
- [x] pinned memory
Examples in `custom_allocations`
HPE seems to be have some cray specific allocations based …
-
In some of my tests (especially of the memory-hard formats on multi-socket/NUMA systems but not only there and not only of those), setting `GOMP_CPU_AFFINITY` to cover the full range of logical CPUs i…
-
### Describe the bug
If the underlying transformers all accept sparse input data, `ColumnTransformer` should also be able to accept sparse input data. That's indeed the case for the `csr`, `csc`, `li…
-
On each iteration of the page rank algorithm we construct a single parallel region as opposed. Since the overhead of spawning teams is very high.
Further. Applied a couple of different optimizatio…
-
### 🐛 Describe the bug
Building wheel torch-2.6.0a0+gitb86b534
-- Building version 2.6.0a0+gitb86b534
cmake -GNinja -DBUILD_PYTHON=True -DBUILD_TEST=True -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_…
-
## Question
Are there examples of Trilinos configure options, and RHEL environment options, for the Tpetra/Kokkos linear algebra stack on a **single MPI rank** with **OpenMP**? The CPU is dual soc…
-
I noticed something similar to the MPI issue with OpenMP since recently: Is it possible that, similar to the recent intel MPI issue, there is another issue that some Python packages started to pull in…