5 ms·

GPU compute in the browser at the speed of native: WebGPU marching cubes

168 points by Twinklebear 2y ago

superkuh 2y ago
Nice. Web browser exploits at the speed of GPU driver bugs. A fount that will never run dry.
- tkzed49 2y ago
  lame. gpus are sick, and the web is cool.
SuboptimalEng 2y ago
Ah, so that's how you do it. Having a template for WebGPU projects is a good idea. I'll have to do the same so I don't waste time setting up web graphics projects. Cool project btw! Adding this to my long list of graphics blogs to read.
bhouston 2y ago
nice! (I notice WebGPU support is now up to somewhere around 60% of all browsers: https://web3dsurvey.com/webgpu https://web3dsurvey.com/webgpu)
dailykoder 2y ago
Nice, now we just need to be able to directly boot into chrome without that OS bloat around it.
codewiz 2y ago
WebGPU has been under development since 2017, and has been a working draft since 2021. What issues are holding the W3C from publishing the final standard? Is there a timeline?
cshenton 2y ago
You absolutely cannot implement stream compaction “at the speed of native” as WebGPU is missing the wave/subgroup intrinsics and globally coherent memory necessary to do that efficiently as possible.
- tehsauce 2y ago
  It's possible you might not need direct access to wave/subgroup ops to implement efficient stream compaction. There's a great old Nvidia blog post on "warp-aggregated atomics" https://developer.nvidia.com/blog/cuda-pro-tip-optimized-filtering-warp-aggregated-atomics/ https://developer.nvidia.com/blog/cuda-pro-tip-optimized-fil... where they show that their compiler is sometimes able to automatically convert global atomic operations into the warp local versions, and achieve the same performance as manually written intrinsics. I was recently curious if 10 years later these same optimizations had made it into other GPUs and platforms besides cuda, so I put together a simple atomics benchmark in WebGPU. https://github.com/PWhiddy/webgpu-atomics-benchmark https://github.com/PWhiddy/webgpu-atomics-benchmark The results seem to indicate that these optimizations are accessible through webgpu on chrome on both MacOS and Linux (with nvidia gpu). Note that I'm not directly testing stream compaction, just incrementing a single global atomic counter. So that would need to be tested to know for sure if the optimization still holds there. If you see any issues with the benchmark or this reasoning please let me know! I am hoping to solidify my knowledge in this area :)
- torginus 2y ago
  I think compilers should be smart enough to substitute group-shared atomics with horizontal ops. If it's not already doing it, it should be! But anyways, Histogram Pyramids is a more efficient algorithm for implementing parallel scan anyways. It essentially builds a series of 3D buffers, each having half the dimension of the previous level, and each value containing the sum of the amounts in each underlying cells, with the top cube being just a single value, the total amount of cells. Then instead of doing the second pass where you figure out what index thread is supposed to write to, and writing it to a buffer, you just simply drill down into said cubes and figure out the index at the invocation of the meshing part by looking at your thread index (lets say 1526), and looking at the 8 smaller cubes (okay, cube 1 has 516 entries, so 1100 to go, cube 2 has 1031 entries, so 69 to go, cube 3 has 225 entries, so we go to cube 3), and recursively repeat until you find the index. Since all threads in a group tend go into the same cubes, all threads tend to read the same bits of memory until getting down to the bottom levels, making it very GPU cache friendly (divergent reads kill GPGPU perf). Forgive me if I got the technical terminology wrong, I haven't actually worked on GPGPU in more than a decade, but it's fun to not that something that I did cca 2011 as an undergrad is suddenly relevant again (in which I implemented HistoPyramids from a 2007ish paper, and Marching Cubes, an 1980s algorithm). Everything old is new again.
lukko 2y ago
Hmm, why does the hydrogen atom look like a d orbital? Isn't it 1s1 in the ground state so the probability distribution would look like a sphere.
- plus 2y ago
  I assume it's just mislabeled, it's a high-angular-momentum hydrogenic orbital, chosen because it looks cool and because it's trivial to evaluate (a spherical harmonic times a simple radial term).
pjmlp 2y ago
Only if we are talking about the state of native in ca 2015.
worik 2y ago
Exciting! But: "Error: Your browser does not support WebGPU" Sigh
- jsheard 2y ago
  Chrome and Chromelikes are still the only browsers shipping stable WebGPU, on Firefox it's behind a flag, and on Safari it's only on the TP branch. Then on Chrome it's not available on Linux yet, only Windows and Android, and only on a subset of Android GPUs. We have a way to go yet.
  - capitainenemo 2y ago
    I was trying it out in nightly firefox, and regardless of the webgpu flags I tried, it still errored: Shader '' parsing error: the type of `SCAN_BLOCK_SIZE` is expected to be `u32`, but got `i32` 10 │ @id(0) override SCAN_BLOCK_SIZE: u32 = 512; │ ^^^^^^^^^^^^^^^ definition of `SCAN_BLOCK_SIZE`
spxneo 2y ago
but somebody told me a while back on here that WebGPU was outdated? whats the consensus ?
- ramon156 2y ago
  Why would wGPU be outdated if it was in the middle of being supported for browsers?
  - spxneo 2y ago
    this is what I want to know but someone said it was not suitable for running GPU powered games on here i will see if i can dig up the thread
    - pjmlp 2y ago
      Depends on how much GB you want to force into users browsers, plan to work around browser blacklists and deep variation in API support, and protect the game code. There is a reason why in 2024, there is yet to exist a WebGL 2.0 game that can match Infinity Blade from 2011, the game used by Apple to demo iPhone's OpenGL ES 3.0 capabilities. WebGPU on top of that, is Chrome only for the time being, still years away from a sound 1.0 release on Safari and Firefox.
    - lukan 2y ago
      It sounds like that someone is really just "someone". And people online have lots of opinions. Not all worth reading. But that someone probably just said, that with WebGPU you do not get the power you would have with a native feature set and this is true. So we likely won't see AAA games anytime soon in the browser. But it is definitely suitable for games in general.
- chrysoprace 2y ago
  You may be thinking of WebGL. My understanding is that WebGPU is effectively supposed to supersede WebGL.
- adfm 2y ago

168 points by Twinklebear 2y ago

superkuh 2y ago

Nice. Web browser exploits at the speed of GPU driver bugs. A fount that will never run dry.

tkzed49 2y ago
lame. gpus are sick, and the web is cool.

SuboptimalEng 2y ago

Ah, so that's how you do it. Having a template for WebGPU projects is a good idea. I'll have to do the same so I don't waste time setting up web graphics projects. Cool project btw! Adding this to my long list of graphics blogs to read.

bhouston 2y ago

nice! (I notice WebGPU support is now up to somewhere around 60% of all browsers: https://web3dsurvey.com/webgpu https://web3dsurvey.com/webgpu)

dailykoder 2y ago

Nice, now we just need to be able to directly boot into chrome without that OS bloat around it.

codewiz 2y ago

WebGPU has been under development since 2017, and has been a working draft since 2021. What issues are holding the W3C from publishing the final standard? Is there a timeline?

cshenton 2y ago

You absolutely cannot implement stream compaction “at the speed of native” as WebGPU is missing the wave/subgroup intrinsics and globally coherent memory necessary to do that efficiently as possible.

tehsauce 2y ago
It's possible you might not need direct access to wave/subgroup ops to implement efficient stream compaction. There's a great old Nvidia blog post on "warp-aggregated atomics" https://developer.nvidia.com/blog/cuda-pro-tip-optimized-filtering-warp-aggregated-atomics/ https://developer.nvidia.com/blog/cuda-pro-tip-optimized-fil... where they show that their compiler is sometimes able to automatically convert global atomic operations into the warp local versions, and achieve the same performance as manually written intrinsics. I was recently curious if 10 years later these same optimizations had made it into other GPUs and platforms besides cuda, so I put together a simple atomics benchmark in WebGPU. https://github.com/PWhiddy/webgpu-atomics-benchmark https://github.com/PWhiddy/webgpu-atomics-benchmark The results seem to indicate that these optimizations are accessible through webgpu on chrome on both MacOS and Linux (with nvidia gpu). Note that I'm not directly testing stream compaction, just incrementing a single global atomic counter. So that would need to be tested to know for sure if the optimization still holds there. If you see any issues with the benchmark or this reasoning please let me know! I am hoping to solidify my knowledge in this area :)
torginus 2y ago
I think compilers should be smart enough to substitute group-shared atomics with horizontal ops. If it's not already doing it, it should be! But anyways, Histogram Pyramids is a more efficient algorithm for implementing parallel scan anyways. It essentially builds a series of 3D buffers, each having half the dimension of the previous level, and each value containing the sum of the amounts in each underlying cells, with the top cube being just a single value, the total amount of cells. Then instead of doing the second pass where you figure out what index thread is supposed to write to, and writing it to a buffer, you just simply drill down into said cubes and figure out the index at the invocation of the meshing part by looking at your thread index (lets say 1526), and looking at the 8 smaller cubes (okay, cube 1 has 516 entries, so 1100 to go, cube 2 has 1031 entries, so 69 to go, cube 3 has 225 entries, so we go to cube 3), and recursively repeat until you find the index. Since all threads in a group tend go into the same cubes, all threads tend to read the same bits of memory until getting down to the bottom levels, making it very GPU cache friendly (divergent reads kill GPGPU perf). Forgive me if I got the technical terminology wrong, I haven't actually worked on GPGPU in more than a decade, but it's fun to not that something that I did cca 2011 as an undergrad is suddenly relevant again (in which I implemented HistoPyramids from a 2007ish paper, and Marching Cubes, an 1980s algorithm). Everything old is new again.

lukko 2y ago

Hmm, why does the hydrogen atom look like a d orbital? Isn't it 1s1 in the ground state so the probability distribution would look like a sphere.

plus 2y ago
I assume it's just mislabeled, it's a high-angular-momentum hydrogenic orbital, chosen because it looks cool and because it's trivial to evaluate (a spherical harmonic times a simple radial term).

pjmlp 2y ago

Only if we are talking about the state of native in ca 2015.

worik 2y ago

Exciting! But: "Error: Your browser does not support WebGPU" Sigh

jsheard 2y ago
Chrome and Chromelikes are still the only browsers shipping stable WebGPU, on Firefox it's behind a flag, and on Safari it's only on the TP branch. Then on Chrome it's not available on Linux yet, only Windows and Android, and only on a subset of Android GPUs. We have a way to go yet.
- capitainenemo 2y ago
  I was trying it out in nightly firefox, and regardless of the webgpu flags I tried, it still errored: Shader '' parsing error: the type of `SCAN_BLOCK_SIZE` is expected to be `u32`, but got `i32` 10 │ @id(0) override SCAN_BLOCK_SIZE: u32 = 512; │ ^^^^^^^^^^^^^^^ definition of `SCAN_BLOCK_SIZE`

spxneo 2y ago

but somebody told me a while back on here that WebGPU was outdated? whats the consensus ?

ramon156 2y ago
Why would wGPU be outdated if it was in the middle of being supported for browsers?
- spxneo 2y ago
  this is what I want to know but someone said it was not suitable for running GPU powered games on here i will see if i can dig up the thread
  - pjmlp 2y ago
    Depends on how much GB you want to force into users browsers, plan to work around browser blacklists and deep variation in API support, and protect the game code. There is a reason why in 2024, there is yet to exist a WebGL 2.0 game that can match Infinity Blade from 2011, the game used by Apple to demo iPhone's OpenGL ES 3.0 capabilities. WebGPU on top of that, is Chrome only for the time being, still years away from a sound 1.0 release on Safari and Firefox.
  - lukan 2y ago
    It sounds like that someone is really just "someone". And people online have lots of opinions. Not all worth reading. But that someone probably just said, that with WebGPU you do not get the power you would have with a native feature set and this is true. So we likely won't see AAA games anytime soon in the browser. But it is definitely suitable for games in general.
chrysoprace 2y ago
You may be thinking of WebGL. My understanding is that WebGPU is effectively supposed to supersede WebGL.
adfm 2y ago