WebGPU Shading Language

Re:WebGPU Shading Language by Jasper_ on Tuesday March 10, 2020 @11:21AM (#22536166)

So this is in a weird place. It's trying to be a floor wax and a desert topping, that is, both a compile target and a human writable language, and the committee is having trouble deciding the vision and goals for the language.

You can see some of that contention here, https://github.com/gpuweb/gpuweb/issues/586 , along with other issues filed.

This is going to be a bumpy ride. I hope it turns out okay.

[ Parent | on HN ]

Re:WebGPU Shading Language by fulafel on Tuesday March 10, 2020 @11:40AM (#22536367)

Based on the current version of the "Goals" text it's more of a SPIR-V surface syntax: "Trivially convertable to SPIR-V"

Sounds like a good thing at first blush, if shaders can be then decompiled to this in debuggers etc.

[ Parent | on HN ]

Re:WebGPU Shading Language by Jasper_ on Tuesday March 10, 2020 @12:02PM (#22536593)

Apple is pushing back on that wording. https://github.com/gpuweb/gpuweb/issues/582

[ Parent | on HN ]

Re:WebGPU Shading Language by kvark on Tuesday March 10, 2020 @04:25PM (#22539310)

That's true, but also a reflection of the whole WebGPU standardization process. It tries to be low-level enough to be performant yet still human writable. I agree it's tricky, but just because it's hard to formulate the concrete goals doesn't mean that there is nothing to discover.

[ Parent | on HN ]