kompute VS SPIRV-Cross

The two I know of are IREE and Kompute[1]. I'm not sure how much momentum the latter has, I don't see it referenced much. There's also a growing body of work that uses Vulkan indirectly through WebGPU. This is currently lagging in performance due to lack of subgroups and cooperative matrix mult, but I see that gap closing. There I think wonnx[2] has the most momentum, but I am aware of other efforts.

[1]: https://kompute.cc/

[2]: https://github.com/webonnx/wonnx

Great to see the positive momentum of this framework! Best wishes and upvotes from the Vulkan Kompute team :)

To a first approximation, Kompute[1] is that. It doesn't seem to be catching on, I'm seeing more buzz around WebGPU solutions, including wonnx[2] and more hand-rolled approaches, and IREE[3], the latter of which has a Vulkan back-end.

[1]: https://kompute.cc/

[2]: https://github.com/webonnx/wonnx

[3]: https://github.com/openxla/iree

I look in an example and see similar instructions, stating that the build should be quite simple. But again, it doesn't work. It generates a bunch of folders with Visual Studio stuff, but no executables, no libraries, or anything like that.

I can't figure out how, and there are no tutorials. According to https://kompute.cc/overview/build-system.html I should simply run "cmake -Bbuild". But this doesn't output what I need, and when I look in the Makefile I get the sense that this is more an example Makefile... but then that contradicts the above tutorial.

Kompute is in my opinion good example to take inspiration for abstractions.

Currently, there's a few Vulkan compute frameworks floating around (like Kompute). I would work with those. Kompute simplifies a lot of the biolerplate and seems like you could benefit from using it.

Try Kompute, a project from the Linux foundation. It is quite simple to use, and does not require deep knowledge of graphics API. It’s a bit painful to setup, but it kinda works well (and I have a project going on on it)

This so much.

This makes alot of sense considering the Linux Foundation is also in charge of Kompute which is likely to be the basis of vendor agnostic GPGPU, and thus the basis of vendor agnostic GPU-based machine learning.

There also exists something like SPIRV-Cross which promises to be able to generate code from the SPIRV intermediate representation into Metal and all versions of GLSL and HLSL. I am not sure really how good it is at this point, but going forward we might start to see more high-level shader languages, that compile to SPRIV and then from there to the myriad of different shader formats different platforms expect.

If you have shaders, I believe you can use SPIRV-Cross to generate GLSL, which you can probably get to pass as OpenCL C with just a bunch of macro tweaks, or at worst some small changes to spv-cross.

Regarding reflection, here is a guide: https://github.com/KhronosGroup/SPIRV-Cross/wiki/Reflection-API-user-guide

I am working on an engine that targets Vulkan and Metal. I'm at the point now where I want to be able to dynamically update my shader at runtime to suit the type of data being sent in for drawing. I am currently using offline compilation for my GLSL (for Vulkan) and MSL (for Metal) shaders. What are your workflows for situations like this? For those using tools like SPIR-V Cross and shaderc, what has your experience been with these tools keeping up to date with the latest features in the specs?

For cross-platform support look at WebGPU and Vulkan (e.g,: [0] [1]. Essentially, you would need to write the func in WGSL or GLSL, HLSL or MSL. Each of these can be cross-compiled to SPIR-V (what Vulkan needs) with cross-compilers such as spirv-cross and naga.

Just for reference, the library I'm using (both for compiling the shaders and for reflection), is SPIRV-Cross by the Khronos Group and here you have the docs for the reflection API. I wanted to check out `glslang` but honestly this one so far has worked like a charm.

Aside from SPIRV-Reflect, if you're using SPIRV-cross to cross compile your shaders, there is also a --reflect arg you can pass which spits out reflection info in JSON format. We already need to cross compile from spirv, so it just removes a tool in the chain to depend on.

For your own engine, use the format your API uses. If you need crossplatformness, there is a new path available. Write your shaders in a language that compiles to SPIR-V (HLSL/GLSL are the most obvious languages), and then use SPIR-V Cross to compile the SPIR-V back to HLSL/GLSL for other API to consume.