Seems that some of the folks don't like this, as its eating up available Bigadvs on equipment that can barely run it. I don't have a problem as long as the WUs are completed by the preferred deadlines.
Same boat as you. As long as they complete a WU, they're fine. I don't get why people are making such a fuss out of it because it's not a "8 thread processor" or whatever.
The official people at Stanford haven't made such a big fuss have they? Otherwise, they would've locked the threads and banned people for talking bout it...
The queue in HFM shows units you've done, not units coming to you.
I'll show you a screenshot of what I'm talking about later when I get home. My first SMP client shows your "normal" queue with finished units, my second one shows nothing but the current one and my third one shows my queue from finished to unfinished ones. Either it's true or I smoked too much of the good stuff
Oddly enough, I think he's right. The last result sent is indeed P6701 R2, C24, G34 (position 6).
I'm pretty sure the 9 slots cycle through so those slots 7, 8, and 9 were the WUs done before the ones in slots 0, 1, and 2 etc. Right now I have a GPU client that is working ona WU in slot 3, with 6 'predicted' WUs in slots 4 to 9.
Simple test would be to pick a GPU or other client that is about to finish, and see if you get the 'predicted WU.'