Patrick Stotko

After taking a quick look at SPIR-V, I believe that it is not easily possible to add it as a backend. stdgpu requires C++14 (I guess it is possible to...

Thanks for the suggestion. A SYCL backend also requires relaxing the dependency to thrust (or adding a SYCL backend to thrust which is more difficult I guess). Since there will...

In #104, the global variables in `memory` have been cleaned up.

In #190, `bitset` and `mutex` have been inlined to achieve GPU architecture independence for the compiled library.

Yes, in principle arbitrary types for the `Key` and `T` types of `stdgpu::unordered_map` container (`stdgpu::unordered_set`likewise) are supported. `int, int` is just one very simple choice in the examples. You can...

This would simplify a lot of code that currently simulates templates with some macro magic, e.g. in [Open3D](https://github.com/isl-org/Open3D/blob/8cc39dd9abc3ca95f171af4647d172052f66d862/cpp/tests/core/ParallelFor.ispc#L190). Is there a chance that we get this awesome feature in the...

I am also observing the mentioned GPU hangs. I submitted a respective pull request (see #56) to fix this issue. Until it gets eventually merged, you can apply and try...

In #391, a resource wrapper is introduced to allow for automatic memory management. Since this is highly related to the copy and resize support, future design decisions involving the wrapper...

In #326, `limits` has been inlined as the declarations became obsolete with C++17.

This is a known limitation. Although the required parallel algorithms from thrust used in stdgpu as well as the intermediate interface in stdgpu all support arbitrary `execution_policy`s (where a CUDA...