Originally Posted by kremtok
So I gather from reading your very detailed results so far that there is no clock-for-clock difference between Sandy Bridge and Nehalem, but the Sandy Bridge will clock much higher at much lower temperatures?
Originally Posted by SniperXX
Thats what I gathered from the reading here. I think its impressive.
Well yes and no.. I agree SB will clock higher at lower temps, but I still think it has an edge over my 930 clock for clock. @4300 on SB and 4280 on the 930, SB had an 11K advantage on regular smp, and still an advantage but only 2.5K on a bigadv.
This comparison is not 100% accurate. The 930 rig has winAFC setting priority to high, and slightly tweaked win 7 install. At this point I have not setup winafc and it's using a normal ultimate version for now. Right now I'm just workin on establishing expected PPD numbers for the 2600k at various clocks. I have captured a 6074 for the SMP portion. Once I'm done this I'll capture a bigadv and do the same.
Update on progress.
I am just finishing up SMP tests a 4.9Ghz
P6074 - 2:07 - 30,296PPD
Past 30K on smp!
This is with vcore set to 1.375 in bios, cpuz reporting 1.392, temps are starting to get warm here, average 75 and high of 78 folding. For now this is the highest I'll be testing at. I seem to be stable folding at 5.0 and 5.1 for very short runs, but temps start to climb real fast. I did not let it run long enough to get any frame times.
The jump from 4.9 to 5.0 and beyond looks like it takes much more vcore on my chip. I could not boot windows with any less than 1.450 set in bios, idle temps still really great 30`c and up to 80% load kept temps acceptable around 70`c.
Working on finding lowest vcore for 4.5 which will be the next bench down to re-run the 6074. I'm thinking I'll do 4.0 and stock 3.4 as well for this and the bigadv runs.
Once I've done all the runs for a particular WU, I'll update the OP with a chart of some sort. Edited by Digigami - 1/8/11 at 3:38pm