Slashdot Mirror
Unity Benchmarks Browser WebGL Performance (unity3d.com)
An anonymous reader writes: Jonas Echterhoff from Unity has posted the latest Unity WebGL benchmark results on the Unity blog. He writes, "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." The benchmark concludes that Firefox 42 64-bit is the fastest, Edge takes second, and Chrome and Safari share third place.
New name same buggy pony show. Nothing can beat Chrome
http://saveie6.com/
Back when Epic first ported Unreal Engine to run in a browser, the only browser it really worked well in was Firefox. Now, Firefox is still the best performer for these kinds of tasks.
Now, commence with the replies explaining why Firefox is a memory hog and not worth using, but it's some kinda weird memory hog that for the past 3-4 years actually uses less memory than Chrome or any other browser.
According to the first charts Unity + Chrome 46 runs faster on a 2.6 GHz i7 + GT 750M MacBook Pro than a 3.3.GHz i7 + GTX 960 Windows desktop. Huh?
Do these numbers prove certain browsers are faster, or that Unity is unevenly optimized?
(Note: Firefox 41 is also faster on the MBP, though that can be attributed to Firefox for Windows being only 32-bit)
How can I believe you when you tell me what I don't want to hear?
Why is it so hard for people to label their f***ing axes!
https://xkcd.com/833/
Yes, this is mostly memory related. Browsers have a tendency to either crash when parsing all the code, or to run out of memory when trying to allocate the memory for the game itself (which emscripten implements using a single big array of bytes in JavaScript). See http://docs.unity3d.com/Manual...
Following open web standards
Which will reportedly soon include compatibility with royalty-free web codecs such as the WebM project's VP9. Yet video in royalty-free formats on Edge will be silent, as the same page states that support for royalty-free audio codecs Vorbis and Opus, used with WebM video, is still "under consideration". And does the only operating system for which Edge is available respect the privacy of web users, or does it by default report every visited URL to Microsoft?
I think the conventional wisdom is that Firefox is a memory hog because after you open a bunch of heavy DHTML documents in tabs and then close them, the memory of the single Firefox process becomes fragmented until you close the last tab. Even a "Minimize memory usage" in about:memory may leave the process a hundred megabytes or more larger than that of a fresh launch. Until Electrolysis becomes the default, the only thing commonly run in a separate process is Flash Player, and that's been on its way out for years in favor of DHTML. Chrome, on the other hand, runs documents in separate processes, which lets it release memory to the operating system more thoroughly once those documents are no longer loaded.
I've found it a little flaky for actual production use so far.
That and laptops with Intel GMA that are stuck on OpenGL 1.4. WebGL requires at least OpenGL 2.0, which wasn't supported in Intel IGPs until HD Graphics. This means Westmere (2010) at bare minimum in desktops and full-size laptops, Sandy Bridge (2011) for any sort of performance, or Bay Trail (Q3 2013) for small or convertible laptops.
If you don't want to buy them, don't buy them. If you want a cross platform browser that runs WebGL fast, then use Firefox.
That's fine on the user end, so long as you have administrator access to install Firefox. But if you are developing a site that relies on WebGL, how can you encourage your users to use Firefox? Or should you accept the browser that the user is already using? If so, then you have to test in every browser, and that includes buying a Mac on which to test in Safari.
What you say is technically correct, in the sense of "You don't have to make a web site" or "You don't have to make your site compatible with Safari." It's just that the operator of a public web site is highly unlikely to rationally want to either stop making a web site or turn away users of OS X's pack-in browser or iOS's only browser engine that isn't run remotely. When you do intentionally give up on Safari users, you get vocal critics like this.