ravil-mobile

Hi everyone, I was testing some batched matrix multiplication kernels using HIP from ROCm 4.5.2, 5.1.0 and 5.1.3 suites. My target hardware is Mi200 i.e., gfx90a. The kernels themselves were...

Dr/cpp

2

C++ implementation of Dynamic Rupture Note, I tag/release of the master branch must be made right before merging with the master

[nvhpc]: disable openmp in input/output for planar wave

1

Looking for a bug fix

Added async copies from GPU-to-CPU for point sources

1

It is an attempt to improve data transfers from GPUs to CPUs while evaluating point sources. The implementation uses GPU streams and OpenMP tasks. In stead of waiting fault pages...

Related to dr/cpp

2

### Rationale `m_dynRup` should be `shared pointer` or `weak pointer` because it is needed in multiple places e.g., evaluation of friction laws, output, etc. ### Description Currently, it is declared...

Refactoring MemoryAllocator.cpp

1

**Describe the problem** Too many `#ifdef` macros. It is difficult to read and maintain the code **Expected behavior** Reduce them to minimum

Added fty to SeisSol as a submodule

4

* reads parameters.par from disk and stores the info in a hash table i.e., YAML::Node * FTY throws an exception if a NAMELIST block is not closed. Example: ``` ......

Related PRs: yateto: https://github.com/SeisSol/yateto/pull/46 Device: https://github.com/SeisSol/Device/pull/21

…mNewStress`