I’m afraid to say I got different [relative] results to those published on the site (the Flex cache as bitmap version was by far the highest FPS on my machine), but I’d have to dig into the source to see how these have been put together…
Link to the results and discussion.
Some personal notes. I tend to think a Flash only (non-Flex framework) example would be a better option to a Flex app. Also the Flex example uses a Timer, he probably should try making use of the ENTER_FRAME event to see what results that gets. Also worth trying the fullscreen-then-back trick the Papervision3D guys are using to get a perf boost. Any other tests out there? Something like this would be interesting 😉
Overall though the performance is great on all the examples so things are heating up 🙂