Andreas Reich Wumpf

Wgpu degenerification performance report

TL;DR:

Compile times are 13-16% shorter if you have several backends (always on Windows), slight but consistent improvements if you use only a single one (default on Mac)
Runtime performance of compute passes gets worse by 10-30% (with a 40% worse outlier for a lot of dispatches in a single pass)
Runtime performance of render passes sometimes 17-34% better and sometimes 10% worse 🤷
Stuck doc-gen issue is fixed!
wgpu-info Windows binary size goes down, Mac binary sizes goes up unless you add more backends

Computational Fluid Dynamics (in graphics mostly)

Disclaimer: I have no professional education in Compulational Fluid Dynamics, these are various notes on what I found/learned when trying to learn a bit on the topic. Some of the notes lack sources as I wrote them up later often just remembering some StackExchange answer. Some things I might have misinterpreted.

Computational Fluid Dynamics (in graphics mostly)
General / Overview
- Overarching resources
General Concepts & Keywords

	<!DOCTYPE html>
	<html lang="en-US">

	<head>
	<meta charset="UTF-8">
	<meta name="viewport" content="width=device-width, initial-scale=1">
	<title>Compiling & linking a large shader</title>
	<script src="main.js" type="module"></script>
	<script id="vertexShader" type="x-shader/x-vertex">#version 300 es


	@group(0) @binding(0)
	var<storage, read> numbers_input: array<u32>;

	@group(0) @binding(1)
	var<storage, read_write> numbers_output: array<u32>;

	var<workgroup> shared_atomic: array<atomic<u32>, 64>;
	var<workgroup> shared_workgroup_memory: array<u32, 64>;

	H264, 25fps:
	."C:\Users\andre\ffmpeg.exe" -i %d.png -c:v libx264 -pix_fmt yuv420p output.mp4

	H264, 60fps:
	."C:\Users\andre\ffmpeg.exe" -framerate 60 -i screenshot%d.png -r 60 -c:v libx264 -pix_fmt yuv420p output.mp4

	H265 (windows can't play it!)
	."C:\Users\andre\ffmpeg.exe" -i %d.png -c:v libx265 -pix_fmt yuv420p output.mp4