Page 2 of 2
Re: 334.89 WHQL drivers
Posted: Wed Mar 12, 2014 12:12 pm
by kiore
So new 335 drivers are out, has anyone tried these yet? I am still stuck with a crippled GTX 680 running beside my GTX780ti.
Re: 334.89 WHQL drivers
Posted: Wed Mar 12, 2014 4:28 pm
by Zagen30
There was one report at OCN that it was an improvement, but a) that was on a core 15 WU and b) later WUs that person did saw no improvement.
Over at EVGA, a couple people tried them out and saw no improvement on 9401 WUs. Someone else there reposted an email chain with an EVGA rep where the rep said this is not a driver issue but an FAH issue, and that a fix should be coming soon (read from bottom to top):
Response Rashid via Email 03/10/2014 10:13 AM
Only application developer will be able to provide more info on that. If the issue was related to driver and was under investigation, I would have notified you once the fix is available from us.
But we worked with the app developer and they mentioned that they will be releasing a patch to fix the issue.
Best Regards,
NVIDIA Customer Care
Customer Jacek via CSS Email 03/10/2014 10:10 AM
Why does installing the previous drivers fix this issue?
Response Rashid via Email 03/10/2014 09:50 AM
Sorry but the issue not due to the drivers. We worked with the application developers on this issue and the upcoming patch should take care of it.
Regards,
Rashid
Customer Jacek via CSS Email 03/10/2014 09:40 AM
What about nvidia? Won’t there be any patches to fix the issue with the current Folding@Home application?
Response Rashid via Email 03/10/2014 09:35 AM
Hello Jacek,
We investigated further on the issue related to Folding@Home and soon there should be patch update from the developers which should help is resolving the issue.
Best Regards,
Rashid
Customer Jacek via CSS Email 03/02/2014 05:25 PM
Sure, keep me updated.
Response Rashid via Email 03/02/2014 04:09 PM
Hello,
Sorry for the troubles but this is currently known to us.
We are actively investigating the issue, in coordination with the application developer. Kindly wait for sometime and once the fix is incorporated in upcoming drivers, I will notify you via email.
Appreciate your patience.
Best Regards,
Rashid
Re: 334.89 WHQL drivers
Posted: Wed Mar 12, 2014 4:32 pm
by kiore
I for one am not convinced Rashid...
Seems more like 'spinning at work' than folding at home..
Re: 334.89 WHQL drivers
Posted: Fri Mar 14, 2014 5:16 am
by Turbo_T
As a test case I have loaded the latest drivers, 335.23, and tested the benchmarks. With 2 GTX 480 FTW hydro copper, my open CL marks went from :
327.23
Explicit-12.206 1ns/day
Implicit-48.0077 ns/day
to
335.23
Expicit-20.228 ns/day
Implicit-88.2339 ns/day
I have had stability problems on one of my cards,the other is rock solid. I am hoping this driver will improve stability as well., more later on that.
Re: 334.89 WHQL drivers
Posted: Fri Mar 14, 2014 8:09 am
by n_w95482
Interesting that there would be such an improvement with Fermi. I'll try it this weekend.
Re: 334.89 WHQL drivers
Posted: Sat Mar 15, 2014 1:22 pm
by kiore
I decided to try the new 335.23 on my i7 rig running GTX 780ti and a GTX 680, shouldn't have bothered
Using PPD as a measure there is no difference between this driver and the other 33* versions for F"H on my rig and my cards on Core 17 P9401's. GTX680 fermi remains crippled at less than 50% of previous output on older drivers.
I had high hopes that this issue would be solved finally but see no evidence of it. My previous 33* driver PPD on these two cards on this series w.u is 780ti 200-250kppd and 680 32-40k currently on 202k and 38k respectively . The 680 was 80-100kppd on older drivers.
Edit PPD measured via HFM.