Search Unity

A bit over a year ago, we released a blog post with performance benchmarks for Unity WebGL, to compare WebGL performance in different browsers. We figured it was time to revisit those benchmarks to see how the numbers have changed.

Microsoft has since released Windows 10 with their new Edge browser (which supports asm.js and is now enabling it by default) – so we were interested to see how that competes. Also, we have an experimental build of Unity using Shared Array Buffers to run multithreaded code, and we wanted to see what kind of performance gains to expect. So we tested this in a nightly build of Firefox with Shared Array Buffer support.

You are welcome try our updated benchmark suite in your own browser here.

Untitled 2

Some notes on methodology changes from last year’s benchmarks:

  • This is using an updated build of the benchmark suite built in Unity 5.3. You can download the project folder for this benchmark suite here, so you can experiment with it locally, or try it on other platforms.
  • This version has been stripped of all of the artwork and eye candy we had in the previous version. They added no value to the benchmarks, and the programmer art used never looked good anyway. Stripping the assets makes the builds smaller, and, more importantly, allows us to redistribute the project folder (see the link above).
  • We omitted the “Mandelbrot GPU” benchmark from the test suite, as that test would really only benchmark the GPU and not show significantly different results on different browsers – so it would skew the overall results to show smaller relative differences.
  • We omitted the comparison to a native standalone build we did last year, as we found those results to be misleading, as we are often running different code on different platforms (like different shader implementations with different quality settings – and also different scripting engines).
  • We tested EdgeHTML 12 (Edge 20.10240.16384.0), which was the latest version when we ran the benchmarks, and which did not have asm.js enabled by default – so we turned it on manually. By now EdgeHTML 13 is released, which has asm.js enabled by default.

Here are the overall scores of the different browsers running the benchmark suite on an Windows 10 machine with a 3.3.GHz i7 CPU and an Nvidia GTX 960 GPU. Firefox 45 with Shared Array Buffers has a greyed out bar, as this is an experimental build of Unity running on a nightly build of Firefox:

Screen Shot 2015-11-30 at 1.51.33 PM

Here are the scores of the different browsers running the benchmark suite on Mac OS X on a Retina MacBook Pro 15” with an 2.6 GHz i7 CPU (so we can see how Safari compares against the other browsers):

Screen Shot 2015-11-10 at 11.12.52 AM

Here are the detailed results for each benchmark on windows (results scaled so that Firefox 41 32-bit equals 1.0 for each test):

Screen Shot 2015-11-30 at 1.52.40 PM

And on OS X (again, scaled so that Firefox equals 1.0 for each test):

Screen Shot 2015-11-10 at 10.00.57 AM

Here are the overall results from the benchmark build we showed last year compared to today, to see how performance has changed since (running on Mac OS X on a Retina MacBook Pro 15” with an 2.6 GHz i7 CPU):

Screen Shot 2015-11-10 at 10.22.29 AM

Finally, here is a benchmark of how long it takes for Unity content to start up. The bars below represent the time in seconds it took between opening the benchmark project and rendering the first frame in different browsers. This is running content from a local disk, so the time to download files from the network is ignored. Firefox performs caching of asm.js compilation results, which means that when the same content is run more than once, compilation can be skipped, and subsequent runs will load faster. So we are showing numbers for both cold (first run) and hot (subsequent run with loading asm.js from cache) runs for Firefox:

Screen Shot 2015-11-30 at 1.35.35 PM

Some findings:

  • Firefox 42 64-bit is currently the fastest shipping browser in most of the benchmarks. The 32-bit version of Firefox is noticeably slower than the 64-bit version.
  • Edge, as a new contender in these benchmarks comes in second, with results close to Firefox (and actually faster the Firefox 32-bit) in most benchmarks. In benchmarks which stress WebGL rendering performance (Particles, Asteroid Field), Edge performs best of all tested browsers.
  • Safari delivers performance comparable to Chrome, up from a year ago when it was significantly slower than Chrome.
  • Internet Explorer 11 is far behind the pack in just about everything, and is too slow to be of much use running Unity WebGL content.
  • Our Unity build with Shared Array Buffers significantly improves performance (by several times in some of the benchmarks). This gives an outlook of future performance gains to expect.
  • Overall, in Firefox, we are seeing an 18% performance increase compared to our results from one year ago. Some of this is attributable to Firefox 41 running faster than Firefox 32, the bigger difference seems to come from improvements in Unity and the emscripten compiler. These results are from OS X, where Firefox has been 64-bit for a long time. On Windows, the improvement since last year will be better, due to the arrival of 64-bit Firefox builds.
  • Looking at load times, most modern browsers are able to load the benchmark project in between 5 and 7 seconds. Firefox is able to cache asm.js compilation, which brings subsequent loads down to 1.5-2 seconds.

38 replies on “Updated WebGL Benchmark Results”

I can’t run the benchmark in Internet Explorer 11 ( Inori version ), the WebGL player return an Out of memory error.

On Linux, without proprietary drivers and running Firefox, the benchmark “runs”, I mean, the whole screen is messed up, but at least it runs. With proprietary drivers on, I got a wonderfull pinkish bubblegum screen, and nothing works.

Sorry about double post, but the WebGL bench just started in Internet Explorer 11 (Inori version), it seems that the Out of Memory is pretty random…

My scores:

Mandelbrot Script: 15 169
Instantiate & Destroy : 10 806
CryptoHast Script: 46 588
Animation & Skinning: 34
Asteroid Field: 2 148
Particles: 13 010
Physics Meshes: 29
Physics Cubes: 38
Physics Spheres: 51
2D Physics Spheres: 103
2D Physics Boxes: 72
AI Agents: 250

Overall Score: 11 476

Glad to see some benchmarks around this. I was especially surprised to see how the different browsers performed on my own hardware. I definitely get different results on Windows 10. I actually get Edge>Firefox>Chrome. Not entirely sure why though.

As far as performance, its getting to “good enough” for many things. I won’t be trying to push AAA content directly in the browser, but for things like social games and apps – its definitely pretty close to where I don’t care so much about the benchmarks anymore.

Is there an available demo built using the shared array buffers? I have a 24-core workstation I’d like to test scalability on.

How about mobile browser?

i tried a small project with just a ugui scroll view, only have 15 fps in release WebGL in iphone 6s safari.

Looking forward to see better performance in mobile browser :D

My results on Firefox 43 64 bits, Windows 8.1 64bits, NVIDIA GTX 780 (358.87):

With Angle (D3D11?) : 67794

With “webgl.disable-angle” : 68138

Seeing your results, either something went wrong in 43 or I’ve done something wrong in my Fox config.

Fedora Linux 23
AMD A8-6600K APU
Firefox 42
54965 overall

The results published in the blog are basically useless, because of being normalized. Since they don’t give the numbers in a way that I can use to compare my own results to the published results, it is all meaningless. Even after you run it yourself, you get no comparative information, and the types of things tested are very different from each other. Presumably there is some sort of typical reason common to published benchmarks why this would be concealed. ;)

OSX 10.11
Firefox 43.0

iMac 24GB Ram, 4GB 780m.

During the Particles benchmark, firefox throws up a dialog “Out of memory. If you are the developer of this content, try allocating more memory to your WebGL build in the WebGL player settings.”

I would also have liked to see results from linux in comparison. I guess we would still see some funny results heading in unknown direction, but knowing is always better than guessing :)

Sad that Firefox 43 and 46 was not tested.
Also It could be a very useful idea to bench also on Linux. The gpu drivers are really different.

And what about the layout engine of the future ? Servo + webrender +browser.html

I took a look at webgl export in 5.3 .. safe to say i won’t be bothering with it, the output scene lighting was completely off from the web standalone, and performance of the scene even after changing settings to be as low as possible was just not worth it.

And the webgl compile time turned my pc into barely useable brick for way too long, no idea who’s bright idea it was to set all compiling processes thread priority onto anything but ‘below normal’ kinda stupid given there is no setting in unity to set the thread priority before it starts, probably the same programmer who set the lighting building processes to the same.

oh and to top it off the webgl version left a 70mb opengl.js with the total size for webgl export being 82mb. Unity provides no way of seeing what the hell its exported out. For comparison the web standalone export was 3mb.

Maybe in 2017 webgl might be worth checking out again right now its bleh, ofc UE current track of releasing decent updates with actual built in engine features as opposed go find and buy such improvements at the asset store might have me switching to that in 2016 instead.

Firefox 43 64 bit / Windows 10 64 / i3 530@4 GHz+8GB RAM / GeForce 750 Ti

My results:

Native WegGL+WebGL2 – 64590
Native WegGL – 61865
Angle D3D11 – 54857
Angle – 50575

Firefox 43.0 Out of memory during particle system.
OS X EL Capitan
Mac Book Pro Retina
mid 2014, 2,5 Ghz, 16 GB RAM,

Great work.

Will Shared Array Buffers be mapped to C# threads so our own multi-threaded code can take advantage of them via IL2CPP?

Any news on progress with SIMD, WebGL 2.0, WebVR and WebAssemby.

I disagree with the decision to omit standalone results from the comparison.

It’s important to understand the performance hit you’re going to take by choosing WebGL over standalone. Which areas are significantly weaker? Which are nearly comparable?

Further, performance is increasing over time on standalone as well and it’s valuable to compare the changes on each platform. If performance increases on standalone are outpacing those of WebGL, that means the gap between the platforms is increasing even in the face of the WebGL improvements and will affect the decision to leverage the platform. Likewise, if the gap is closing, that makes a stronger case for WebGL.

I have to wonder if the comparison was omitted because it potentially paints WebGL in a more negative light, but imho the comparison is crucial to making an informed decision. Sure, we can do the benchmarks ourselves, but why omit the information when you’re already publishing results publicly?

Awesome, I was in the middle of writing a presentation on Unity’s WebGL benchmarks when you updated this today.

Thanks for keeping us in the loop, looking forward to seeing how browsers improve perf over time.

Great to see more info on WebGL builds.

What about mobile ( other WebGL tech seem to work ok, eg Blend4Web, PlayCanvas etc ), would be cool to have WebGL as a potential option for smaller games across desktop and mobile.

Also would be cool to have report on reduced file size ( if any ), of say a single cube, to get an idea of the overhead pre adding assets. Maybe build times as well.

Keep up the great work in this area!
Mal

Comments are closed.