This stuff doesn't look simple to use. I need to read a paper or two on it. It's kind of like what we do in database but at a code level; not a database level.
My gut feel is that Intel is going to add this to help auto-parallelization in their compiler suite. I'd guess that they will have pragmas or some other language mechanism to allow the developer to specify sections too. It should be really interesting to see where they can get performance gains. I would guess that this is one way to improve performance on single-threaded programs without any programming.
Market Data 
Markets 
AI
