Jan Ozer’s Flash Player vs. HTML5 Video Performance Tests

Ozer draws the conclusion that Flash Player’s access to hardware acceleration is the key advantage to its superior performance on Windows. And, indeed, the best results in the whole test were for Flash Player 10.1 on Windows. (I wouldn’t be surprised if Mac OS X eventually offers similar APIs — seems like a serious performance win.)

But there are a lot of other interesting numbers in Ozer’s results. Particularly if you look at Flash Player 10.0, which doesn’t use hardware acceleration on Windows, either. In both Chrome and Firefox, Flash uses about twice as much CPU time to render the same video on the Mac as on Windows. Flash performance is noticeably better in Safari on the Mac than it is in Chrome or Firefox — I did not know that. Video performance in Chrome for Mac — both HTML5 and Flash — is downright terrible. (YouTube ought to stop telling Mac-using Safari users to “Try YouTube in a fast, new web browser!” with a link to Chrome.)

Bottom line: Flash plays H.264 video at least twice as efficiently on Windows as on the Mac; Safari’s native HTML5 video playback is very efficient.

The whole test might need to be taken with a grain of salt though. Ozer couldn’t get Bootcamp working on his MacBook Pro, so:

I tested on the Mac using a MacBook Pro (3.06 GHz Core 2 Duo, 8 GB RAM, OS 10.6.2) while testing on Windows using an Hewlett Packard 8710w mobile workstation (2.2 GHz Core 2 Duo system running 64-bit Windows 7 with 2 GB of RAM).

Seems to me the Mac hardware was significantly faster than the Windows hardware — so I suspect his results are misleading with regard to just how much more efficient Flash is on Windows than Mac OS X.

Thursday, 11 March 2010