Commits · d5c5fac4f6cd2007c3b8236897aafc002b6ea705 · waLBerla / waLBerla

29 Jun, 2021 1 commit
- Revamp LB GPU Benchmark App for In-Place Streaming · d5c5fac4
  Frederik Hennig authored Jun 29, 2021 and Markus Holzer committed Jun 29, 2021
  
  d5c5fac4
12 Apr, 2021 1 commit
- FIX: SP/DP conversion warnings · 05b33755
  Markus Holzer authored Apr 12, 2021 and Helen Schottenhamml committed Apr 12, 2021
  
  05b33755
10 Jun, 2020 1 commit

modernize cuda integration; enable cuda + codegen tests · 04d183bf

Dominik Thoennes authored Jun 10, 2020

use cuda as a first level language in CMake with `enable_language(CUDA)`
enable tests where cuda and codegen are both activated
various minor fixes

04d183bf

04 Dec, 2019 1 commit
- Cleaned up UniformGridGPU benchmark and adapted it to changes in python codegen packages · 550358c0
  Martin Bauer authored Dec 04, 2019
  
  550358c0
25 Jun, 2019 1 commit
- UniformGridGPU more configurations · e33c19be
  Martin Bauer authored Jun 25, 2019
  
  e33c19be
24 Jun, 2019 1 commit
- GPU benchmark: added LB storage pattern to python callback · 2ad061e8
  Martin Bauer authored Jun 24, 2019
  
  2ad061e8
18 Jun, 2019 1 commit

Extensions in UniformGridGPU benchmark · 58b3e682

Martin Bauer authored Jun 18, 2019

- different stencils supported
- mpi datatype communication: full vs. required pdfs only
- adapted config file

58b3e682

14 Jun, 2019 1 commit
- UniformGridGPU: New app with AA pattern (single PDF field) · 6a61fc77
  Martin Bauer authored Jun 14, 2019
  
  6a61fc77
13 Jun, 2019 1 commit
- Generalized UniformGridGPU benchmark · 0b75cce0
  Martin Bauer authored Jun 13, 2019
  
  0b75cce0
25 Apr, 2019 8 commits
- UniformGridGPU: select comm type by string instead of number · 6d6abeec
  Martin Bauer authored Apr 25, 2019
  
  6d6abeec
- UniformGridGPU small changes · ad9c95d5
  Martin Bauer authored Apr 25, 2019
```
- MPI barrier between runs
- bugfix in overlap_benchmark config
```
  ad9c95d5
- UniformGridGPU improvements · 8199280e
  Martin Bauer authored Apr 25, 2019
```
- more overlap strategies
- inner/outer split width is configurable
```
  8199280e
- NVTX support · c2609c18
  Martin Bauer authored Apr 24, 2019
  
  c2609c18
- UniformGridGPU · 6dc55108
  Martin Bauer authored Apr 24, 2019
```
- fix in time measurement
- options to disable boundaries and/or communication
- priority for communication & outer part of kernel
```
  6dc55108
- UniformGridGPU: Fix in timing for shorter runs · 628f2060
  Martin Bauer authored Apr 18, 2019
  
  628f2060
- Extended UniformGridGPU simulation setup · 79668534
  Martin Bauer authored Apr 18, 2019
  
  79668534
- GPU communication · 330c43f6
  João Victor Tozatti Risso authored Jan 27, 2019 and Martin Bauer committed Apr 25, 2019
```
- parameter files for UniformGridGPU benchmark
- memcpy-based GPU PackInfo
```
  330c43f6
22 Jan, 2019 3 commits
- UniformGridGPU: Changes related to intel compiler · 997ec0b7
  Martin Bauer authored Nov 15, 2018
```
- intel 19 internal bug when using auto, and implicit capture in lambdas
```
  997ec0b7
- Function for CUDA device selection based on MPI Rank · e17f9a53
  Martin Bauer authored Nov 15, 2018
  
  e17f9a53
- CUDA Parallel Streams & UniformGridGPU Benchmark · a0c86311
  Martin Bauer authored Oct 31, 2018
```
- LB on GPU, Uniform Grid Benchmark app
- helper class to schedule tasks to multiple CUDA streams
```
  a0c86311