alpaka

alpaka copied to clipboard

Memcpy between the same buffer

3

#28 proposes a mechanism to perform a zero copy between two buffers. During the implementation I wondered what the behavior of the standard `alpaka::memcpy` is when destination and source buffer...

bernhardmgruber

CMake target for each accelerator

9

Based on [this](https://github.com/alpaka-group/alpaka/pull/942#issuecomment-593970673) I suggest that we add CMake target for each accelerator backend. Currently alpaka CMake behaves to optimal fits the CI workflow and is activating all available backends....

psychocoderHPC

CMake argument documentation in Sphinx Doc

3

I think a documentation about the different CMake arguments in the Sphinx documentation would be really helpful. For example, today I had some difficulties compiling my Alpaka code with the...

SimeonEhrig

Add tests and example for multi GPU usage

2

Inspecting alpaka's implementations when thinking about zero-copying as part of #1820 I wondered whether alpaka actually supports copying buffers between two GPUs on e.g. the CUDA backend. Searching alpaka for...

bernhardmgruber

[RFC] Visual Studio requires CMake 3.24

3

While working on #1890 I discovered that the Visual Studio build instructions generated by CMake can only consume `SYSTEM` includes since CMake 3.24. I attribute it to simple luck that...

j-stephan

Make CUDA extended lambdas required

1

Alpaka has an option (`alpaka_CUDA_EXPT_EXTENDED_LAMBDA`) to turn on CUDA's extended lambdas, which allow the use of lambda functions inside and as kernels. This option can be turned off and we...

bernhardmgruber

We need a small script which is emulating the CI on the local system. Start the container prepare all the same way the CI is preparing the environment and run...

psychocoderHPC

As @fwyzard [suggested](https://github.com/alpaka-group/alpaka/pull/1712#issuecomment-1330549436) a while back, we should have a primitive in alpaka that helps writing grid strided loops. For 1D index spaces, a range construct like `elements_with_stride` could be...

bernhardmgruber

We merged the Babelstream benchmark in https://github.com/alpaka-group/alpaka/pull/1846. The benchmark is mainly written with GPUs in mind, but since alpaka supports many backends, we should also try to account for the...

bernhardmgruber

[Discussion] Improving the Doxygen API

8

During the workshop today, there was question about the best practice of using the Doxygen documentation. In my opinion it's hard to read the documentation, in special if you are...

SimeonEhrig