-
As mentioned https://github.com/alpaka-group/alpaka/pull/1549#issuecomment-998748452 the openAcc runtime test is showing illegal memory access, the reason is not clear:
[linux_nvhpc-21.9_cuda-11.4_…
-
This is both an issue and a question.
## Question: How does the `atomic` test pass for the CUDA/HIP backend?
The atomic test tests 8-bit atomic operations:
https://github.com/alpaka-group/alpak…
-
The `ExecCpuOmp4` / `AccCpuOmp4` does not compile with ICC because: `... called in "omp target" region must appear in prior "omp declare target" pragma`.
This affects at least `core::mapIdx`, `block::…
-
We have tested ring model with NMODL's OpenACC backend but not BBP model. The build script on BB5 to use NMODL:
```bash
CUR_DIR=`pwd`
module load cmake gcc/6.4.0 flex/2.6.3 bison/3.0.5 python-d…
-
### Describe the bug
rocBLAS or Tensile will not correctly generate .co and .dat files when building. (TensileCreateLibraryFiles)
### To Reproduce
7294a708d91a68e6179d6d1bc74b254686ee8030 (Also l…
-
Hi,
I would like to ask your advice on improving the performance of a small Kokkos kernel on the CPU side. The kernel is an expensive diffusion routine from NEPTUNE, a spectral-element weather mode…
-
I believe that it would be useful to test the `NEURON/CoreNEURON` integration tests with the `NMODL` `OpenACC` backend in `GPU`s to avoid any regressions.
This can be done in the `Gitlab CI`
-
This compiler supports OpenCL?
-
**When compiling from source for Turing GPU (RTX 2080 Ti), nvfortran throws a fatal error:**
`FAILED: lib_xtb.a.p/src_solv_ddcosmo_solver.f90.o lib_xtb.a.p/xtb_solv_ddcosmo_solver.mod
nvfortran -I…
-
This issue was raised when running `nrnivmodl-core` with the following `ModelDB` model https://senselab.med.yale.edu/ModelDB/ShowModel?model=19176&file=%2fHCN2k%2fhcn2.mod#tabs-2 using `PGI 19.4` and …