alex/vello
1
0
Fork 0
mirror of https://github.com/italicsjenga/vello.git synced 2025-01-08 20:01:30 +11:00
Code Issues Packages Projects Releases Wiki Activity

typo fix

Browse source 
Co-authored-by: Vincent Isambart <vincent.isambart@gmail.com>
This commit is contained in:
Raph Levien 2023-01-06 20:15:41 -08:00 committed by GitHub
parent 767997e80a
commit 96a64f8abb
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 1 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
doc/roadmap_2023.md
View file

@ -108,7 +108,7 @@ We're now doing all the shader implementation work in WGSL and using wgpu as the
One issue is that Vello becomes a fairly heavy dependency, as it drags along an entire WebGPU implementation, especially runtime shader compilation. For some applications, that may not be an issue (and not at all if they're already using wgpu), but for others it might be.
Another issue is the overhead of WebGPU compared to native, which we hope is not very big but we haven't quantified carefully. Related, when we get into extensions that are not yet available in WebGPU, those could be implemented fairly straightforwadly on native. On the other hand, another path to those extensions is to prototype them within the WebGPU framework, and that has the benefit of providing useful experience to the WebGPU standardization process. The extensions we care about most are subgroups, device-scoped barriers (to unlock single-pass prefix scans), and descriptor indexing (as described above to reduce buffer allocation and copying for images).
Another issue is the overhead of WebGPU compared to native, which we hope is not very big but we haven't quantified carefully. Related, when we get into extensions that are not yet available in WebGPU, those could be implemented fairly straightforwardly on native. On the other hand, another path to those extensions is to prototype them within the WebGPU framework, and that has the benefit of providing useful experience to the WebGPU standardization process. The extensions we care about most are subgroups, device-scoped barriers (to unlock single-pass prefix scans), and descriptor indexing (as described above to reduce buffer allocation and copying for images).
Vello now has a "recording" abstraction that includes lightweight proxies for resources such as buffers and textures, and a declarative approach to specifying the graph of compute shader dispatches. This abstraction is an appealing alternative to an object-oriented hardware abstraction layer (as was the case in the pre-WSGL version of piet-gpu). We also think this abstraction could support focused, lightweight back-ends for more native GPU APIs. The relative priority of building such back-ends is not clear, but we did want a design where we could gain back some of the performance that was given up in the move to WebGPU.