I've posed this a few places but here are my observations of ML & FCPx:

Ok so I'm not running a 100 core cluster just a dual core 2012 i7 mac mini with a AMD Radeon HD 6630M / 256 MB but wow, FCPx is a lot faster on the transcoding side. I swear I benchmarked 3x for 1080p60 transcoding now it's more like 1x (one min of footage take one min to transcode to prores proxy and regular prores)!!!

Also OSX finally plays AVCHD natively even my nice 1080p60 footage at 28mbps does not stutter and it uses way way less CPU cycles.

Looks like they are finally using Open GL. (On my shite box Windows 7 laptops I've been able to play native avchd with low cpu usage for the past year so well done apple for catching up)

This is great new for us cheapskates using minis instead of Mac Pros

Can someone confirm this speed increase just in case I've been drinking too much of the Job's Kool-Aid.