Preface summary: Recently purchased two 970 FTWs to replace my two GTX 680's (4GB VRAM models) in my main PC. I've been recording my own folding statistics since the addition of the 970's. Initially I was only running the 970's in the main PC, and relegated one 680 to the secondary PC (only one PCIE slot - the specifications for both PCs are in the signature line), in an attempt to pull all "like" WUs to compare the 970 to the 680.
 
However, as of today, I decided to go ahead and utilize the idle 680, that was not being used to produce work, to put in the main PC to see if I could capture identical WUs. What do you know....
 
I've been waiting for the 'apple-to-apple' comparison event to occur before sharing findings, and go figure, after starting the [link=mailto:F@H]F@H[/link] Client, I pulled all 9201 (Core 17) WUs. Obviously we're aware that Core 18 hasn't been fully utilizing (for lack of better term) the 900 series, so this isn't the greatest comparison for future reference, but for today, I'd suspect that it is. What was reported by FAHControl is shown below (excerpt from log + screen shot):
 
----------
25-Nov-14 - ADDED 680 TO MAIN RIG (Case side off & Room Window Open)
 
WU1: CPU0 - 3930k (4400MHz) - TDP Watts: ~162 - Project: 9011 (Core A4) - TPF: 1:06 - PPD: 26222
WU2: GPU0 - 680 4GB - (1097 MHz core [STOCK]) - TDP: 55% - Temp: 55C - Fan: 49% - Project: 9201 (Core 17) - TPF: 1:58 - PPD: 249879 (PCIE Slot 3 [Bottom])
WU3: GPU1 - 970 FTW - (1490 MHz core [+60]) - TDP: ~73-78% - Temp: ~66c - Fan: 82% - Project: 9201 (Core 17) - TPF: 1:51 - PPD: 259566 (PCIE Slot 2 [Middle])
WU4: GPU2 - 970 FTW - (1480 MHz core [+75]) - TDP: ~71-73% - Temp: ~60c - Fan: 70% - Project: 9201 (Core 17) - TPF: 1:56 - PPD: 259337 (PCIE Slot 1 [Top])
 
Total Watts: ~704-712w
Peak Estimated PPD: 796424
----------
 

 
As a side note, this has quite honestly surpassed all previous production from any configuration I've owned. Quite pleased today.
 
- Mike

Would not have expected a 680 to be that high... that's outproducing most 780 users I see reporting.

Not really sure what to say, other than I agree with everyone's surprise reaction. I did not expect the TPF to be soo close to the 970 either.

See here: http://forums.evga.com/FindPost/2258148
 
Everyone has my apologies. Please disregard this thread (it should probably be deleted in order to not mis-guide any future readers.)
 
What still confuses me however, is why the 9201 WUs still render nearly the same TPF for the first 10%-or-so percentage points for all cards? (The same TPFs were seen again today, on all 9201 WUs until reaching about 10% completion on all folding slots.)

@yodap: Drivers for the 970s are the most current 344.75. However, upon the addition of the 680, the 680 associated itself with 344.48 automatically. (The extracted driver files already resided on my system.)
 
Anyone have any insight regarding any possibilities why the 680 reports near-identical TPFs for a given WU until roughly 10% completion?