yuzu/yuzu-emu.github.io
1
0
Fork 0
mirror of https://github.com/yuzu-emu/yuzu-emu.github.io.git synced 2025-07-27 12:11:56 +00:00
Code Issues Packages Projects Releases Wiki Activity

Update site/content/entry/yuzu-progress-report-feb-2021/index.md

Browse source 
Co-authored-by: Ameer J <52414509+ameerj@users.noreply.github.com>
This commit is contained in:
Honghoa 2021-03-10 16:42:58 -03:00 committed by GitHub
parent 3c9c60cbdf
commit 2ad081ecac
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
site/content/entry/yuzu-progress-report-feb-2021/index.md
View file

@ -79,7 +79,7 @@ Thus, these programs were promptly named as [shaders](https://en.wikipedia.org/w
Modern GPUs are designed to break down their workload into smaller sized problems, which in turn are processed simultaneously in the many compute units of the card (entities akin to cores in a CPU).
The reason for this design choice is simply because parallelisation is a very efficient scheme to process computer graphics. A single instruction is capable of operating over many components of data at the same time, such as the vertices and textures of a 3D scene, and produce separate results for each parallel thread of execution, which is generally a pixel. This increases the throughput of information, especially when compared against the performance of running the same operations on a CPU.
Their potential isn't limited to just these functions though.
It is possible to write programs that won't necessarily operate over graphics, yet still benefit from these characteristics.
It is possible to write programs that won't necessarily operate over graphics, but can still take advantage of the high levels of parallelisation provided by GPUs.
This is known as `GPGPU` - [General-purpose computing on graphics processing units](https://en.wikipedia.org/wiki/General-purpose_computing_on_graphics_processing_units) - and it's intended to be used when there is a problem that can be separated into a number of parallel tasks in order to be processed more efficiently. These problems are commonly called `embarrassingly parallel problems`.
One of these cases was the [use of compute shaders to decode ASTC textures](https://github.com/yuzu-emu/yuzu/pull/5927).