Ran 1.8 yesterday and I’d like to say thanks for beefing up CPU rendering. I have a hyper threading 4 core i7 (8 virtual cores, it feels like these are being leveraged now?!) with 12 GB of RAM and I was having a lot of problems running CPU renders before. I now have much of the functionality I was originally hoping for, albeit a bit later than I would have preferred.
Side note: If anyone out there developing algorithms is interested in some really interesting, exotic stuff (and has some understanding of parallel computing) they should check out the Data Vortex. They have a game changing solution to the problem of high-node-count machines being interconnect limited when packet size is small.