jiangzhaoming

wgsl: Is/should initializer of module-scope variables limited to creation-time expressions?

6

(By saying initializer of module-scope variables, it is actually initializer of `private` variables, since other types of module-scope variables must not have an initializer.) In current WGSL spec, module scope...

Avoid indeterminate value result for `tanh` with large input (f32 88.7 or f16 11.09)

1

Currently the [accuracy requirement of built-in `tanh` ]( https://gpuweb.github.io/gpuweb/wgsl/#floating-point-accuracy:~:text=tanh(x),x ) is defined as `Inherited from sinh(x) / cosh(x)`, while both `sinh(x)` and `cosh(x)` inherited from `exp(x)` and `exp(-x)`. `exp(x)` is...

[JS/WebGPU] Creating devices with subgroup features enabled if possible

4

This CL make WebGPU backend support subgroup features and thus allow using subgroup optimizations in the future. ### Description With this CL WebGPU backends will create devices with subgroups and...

Proposal: expose GPUAdatperInfo from GPUDevice

3

This PR is a proposal to enable accessing the GPUAdapterInfo of the physical adapter that create a GPUDevice by using GPUDevice.adapterInfo. Issue: #4810

Add `subgroups-f16` feature dependency validation to subgroups proposal?

3

Currently in the subgroups proposal we have two WebGPU features, `subgroups` and `subgroups-f16`, to enable corresponding WGSL extensions `subgroups` and `subgroups_f16`. Since WebGPU feature `subgroups-f16` is to enable using WGSL...

Allow getting adapter information from GPUDevice

7

In some situation like third-party ML/render library, we may want to take a `GPUDevice` from user application and work on it, instead of creating it from a `GPUAdapter` under library's...

Subgroups: clarify subgroups operations using f16

3

This PR clarify that using f16 in subgroups operation only requires `f16` WGSL extension enabled together with `subgroups` (since `subgroups_f16` is deprecated), and an adapter that support `shader-f16` GPU feature...