Folding Forum

Hello,
I got a little problem and waiting for your help. i begin to fold again with gtx970 however i could not fix 60kppd, at core 17 it is 60k ppd at core 15 it is 74k ppd ( also there is a serious coil whine ) what might be the problem ?

When i calculated over from the simulation list Project 9201 it should be 285k ppd...

I tried drivers at win 7 ( 64 bit ) 344.60 and 344.65 i could not get a clean installation, i installed the Fah Client several times at different directories but the result is still same.

I dont want to re-install windows again can you please help me ?
http://s17.postimg.org/d7hhncej3/fold1.jpg
http://s2.postimg.org/yvjo2vo7d/fold2.jpg

Mod edit: Changed img links to URL's

Please post your log file as described in this topic - viewtopic.php?f=61&t=26036. Post at least the beginning section with version information, system info and the folding client configuration through the first several percent of a WU being processed.

As you get 8000 points I believe you don't have a passkey requested or installed or not yet enough WU returned under that key (10 WU required)

See here: http://folding.stanford.edu/home/faq/faq-passkey/

Coil whine can happen to any card or PSU, you can try backing off the overclock, finding the coil causing the problem and try and dampen the whine - not sure how but people do do it , google is your friend for that one or try an RMA, I know that some PSU companies will honour RMA for coil whine - not sure about video card vendors.

P7623 has no QRB and 285k PPD for P9201 indicate that Anton uses a passkey.

Core 0x15 has a significant higher power consumption than core 0x17 on Maxwell.
A high OC setting might cause errors that increase folding time and reduce PPD.
Please post the log for that P7623 project.

I don´t know PPD values for P7623 and GTX 970 but 45k PPD is slightly higher than a GTX 680 with high OC settings.
A GTX 980 with high OC settings gets 89k PPD. But it should only be 1.2 times faster an GTX 970, so 45k PPD is quite low.

EDIT: Core 0x15 with 74k PPD should be ok for a GTX 970 but your screenshot shows only 45k PPD.

Heiko

heikosch wrote:P7623 has no QRB and 285k PPD for P9201 indicate that Anton uses a passkey.

From what I read his sentence the 285kPPD is a value he determined by calculation himself; not by the system. Plus the second image in the URL shows 14093 points for the core15 project (no QRB) and exact 8000 for core17 (base point for 9201). With no long breaks in between it follows the " little bit above 3 hours runtime" pattern I have with mine. I still bet: issue with passkey. Log file as requested by Joe_H will get clarity.

First of all i wanted to thank everyone that helped me, i solved the problem it was the passkey..

Coil Whine still happens before the OC so backing off wont work, as i searched it is a general problem that GTX970 users suffers..