Skip to content

Commit

Permalink
Update README.md
Browse files Browse the repository at this point in the history
  • Loading branch information
@SinclaM
SinclaM committed Jan 1, 2024
1 parent 82fc927 commit a2e115b
Show file tree
Hide file tree
Showing 2 changed files with 22 additions and 23 deletions.
43 changes: 21 additions & 22 deletions README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -82,7 +82,8 @@ zig build -Doptimize=ReleaseFast

To target the web (populating `www/` with the all the site's files):
```bash
zig build -Dtarget=wasm32-freestanding -Doptimize=ReleaseFast -Dcpu=generic+bulk_memory+atomics+mutable_globals
zig build --sysroot ~/emsdk/upstream/emscripten -Dtarget=wasm32-emscripten -Doptimize=ReleaseFast
&& sed -i'' -e 's/_emscripten_return_address,/() => {},/g' www/ray-tracer-challenge.js
```

## Performance profiling
Expand All @@ -103,18 +104,13 @@ The ray tracer currently runs about 2x slower on WebAssembly than on native, whi

I also use [hyperfine](https://github.com/sharkdp/hyperfine) for benchmarking.

Performance on the website for complicated scenes can be seriously inhibited by memory consumption, since each web worker
maintains its own complete copy of the scene information. It would be easy to simple share memory, using a `SharedArrayBuffer`
for the WASM memory, but unfortunately Zig currently provides no support (or documentation) for this. This problem becomes
amplified on Chrome, which overestimates the amount of web workers that can run in parallel on my system, leading to
even more inflated memory use and worse performance. At some point I'd like to add a slider for the number of web workers
to use instead of just relying on `navigator.hardwareConcurrency`, which would help here.

## Benchmarks

Below are some benchmarks for scenes that can be found on the website. These benchmarks are not rigorously controlled
and averaged, but rather a general overview of speeds for various scenes. The best way to get a feel for the performance
is to try things out yourself!
and averaged, but rather a general overview of speeds for various scenes. They may also change depending significantly
between Zig compiler versions. For example, I noticed a perfromance regression of up to 25% going from 0.11.0 to the
WIP 0.12.0 (perhaps related to [this similar issue](https://github.com/ziglang/zig/issues/17768)). The best way to
get a feel for the performance is to try things out yourself!

All benchmarks were done on a 2019 MacBook Pro (2.6Ghz, 6-Core Intel i7; 16GB RAM; macOS 12.6.7). WASM specific benchmarks
were done on Firefox 117 using 6 web workers (the maximum number of web workers Firefox will run in parallel, even on my
Expand All @@ -125,20 +121,23 @@ and bounding boxes already made, textures already loaded, etc.), which is suppor
camera movememnt. This preheating is irrelevant for simple scenes, but gives massive speedups for scenes that load
textures or construct BVHs.

Also note that renders on the website are periodically polled for completion. Renders may actually
complete up to 100ms before the reported time, which affects the benchmarks for very short renders.

| Scene | Resolution | Native | WASM | WASM Preheated |
| ------------------------- | -------------- | ----------- | ---------- | ------------------ |
| Cover Scene | 1280x1280 | 1.446 s | 2.313 s | 2.305 s |
| Cubes | 600x300 | 0.190 s | 0.413 s | 0.404 s |
| Cylinders | 800x400 | 0.088 s | 0.106 s | 0.110 s |
| Reflection and Refraction | 400x200 | 0.088 s | 0.161 s | 0.164 s |
| Fresnel | 600x600 | 0.243 s | 0.355 s | 0.341 s |
| Groups | 600x200 | 0.074 s | 0.145 s | 0.148 s |
| Teapot | 250x150 | 0.164 s | 0.364 s | 0.232 s |
| Dragons | 500x200 | 6.991 s | 17.011 s | 4.324 s |
| Nefertiti | 300x500 | 4.630 s | 10.578 s | 7.039 s |
| Earth | 800x400 | 0.082 s | 0.359 s | 0.043 s |
| Skybox[^1] | 800x400 | 1.775 s | 15.031 s | 0.055 s |
| Raytracer REPL Default | 1280x720 | 0.138 s | 0.142 s | 0.140 s |
| Cover Scene | 1280x1280 | 1.413 s | 2.408 s | 2.299 s |
| Cubes | 600x300 | 0.225 s | 0.418 s | 0.407 s |
| Cylinders | 800x400 | 0.111 s | 0.221 s | 0.109 s |
| Reflection and Refraction | 400x200 | 0.113 s | 0.213 s | 0.205 s |
| Fresnel | 600x600 | 0.283 s | 0.429 s | 0.411 s |
| Groups | 600x200 | 0.091 s | 0.217 s | 0.202 s |
| Teapot | 250x150 | 0.175 s | 0.413 s | 0.210 s |
| Dragons | 500x200 | 6.957 s | 12.663 s | 2.492 s |
| Nefertiti | 300x500 | 4.827 s | 6.358 s | 3.036 s |
| Earth | 800x400 | 0.095 s | 0.212 s | 0.103 s |
| Skybox[^1] | 800x400 | 1.466 s | 1.531 s | 0.102 s |
| Raytracer REPL Default | 1280x720 | 0.210 s | 0.220 s | 0.209 s |

## Other implementations
There are many great implementations of the Ray Tracer Challenge. At many points throughout the project, I referred to others
Expand Down
2 changes: 1 addition & 1 deletion www/index.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -22,9 +22,9 @@
</div>
<div id="scene-choices" class="select">
<select id="standard-select">
<option value="cylinders.json">Cylinders</option>
<option value="cover.json">Cover Scene</option>
<option value="cubes.json">Cubes</option>
<option value="cylinders.json">Cylinders</option>
<option value="reflection_and_refraction.json">
Reflection and Refraction
</option>
Expand Down

0 comments on commit a2e115b

Please sign in to comment.