Has anyone tried this out and noticed a nice speed bump? I think there will be a new weekly build out tomorrow so I might have to give it a look.
Ooh, nice. I’ll try tomorrow. 
A few things that crossed my mind, looking at the numbers of that benchmark:
- according to those numbers, the 6.0 client VM is exactly as fast as the 5.0 server VM. Coincidence? ???
- the 6.0 server VM is only 5% faster than the client VM. Makes you wonder what the point of the server VM is now. I’d choose a lower footprint over a 5% better performance for server apps anytime…
It appears that the number in bold above each array of tests is the average of the scores in each test. This is true for all tests (okay, actually I didn’t check all of them) except the Mustang b59 tests. Whoever wrote the table must have mis-pasted the data.