Folding@Home: THGC Needs You -Team 40051

Page 434 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.


I've set mine up to just advanced now. Thanks.
 


Don't forget the boost speed will only register under load, i.e. while folding, sounds like the 409mhz you are seeing is when the card is idle and is running power saving mode.

So 1367 + 20 your load/boost speed should now register 1387mhz, mines at 1399mhz, below.

eCYkgIV.png
 


As I have no upgrades planned until Pascal at the earliest I know it's only a matter of time, it's been nice up here though and the view is lovely! :lol:
 
This is the problem i'm having. My GPU clock has been at 405 MHz consistently for days now, and my PPD has dropped. I can't figure out what I'm doing wrong...

...and Happy New Year.

Capture_zpsn5lo96kt.png
[/URL][/img]
 


+1

just encountered this myself last night when trying out my overclock with F@H bench/ Valley with the gpu getting stuck at that 4xx MHz after a driver crash.

Restart the computer and you'll be all good to go.
 
My load at 405 MHz was 95%. I'm back to my default speed now, Gpu power is 86% and load is 95% Do I need to reboot everytime I change the boost clock? And I thought that when I did change the boost clock that I was changing it by percent -- 10%, 20%, etc., but on the dial it looked like it only moved units. Even then, the digital readout doesn't change.

 
Well that screenshot is showing your gpu running @1404mhz without any core increase?!?

When you move the core clock slider it goes up in MHz, so +5mhz +10 MHz etc etc, it should instantly change the boost speed of the gpu as soon as you apply it, press the tick button.

I thought you said your boost was 1367mhz?
 
My 980ti has a default boost of 1354 but while folding certain wu's as long as the heat stays below 64C it will run at 1366mhz.

I now have it clocked @1399mhz with the fan profile adjusted to stop it from thermal throttling.

The 980 seems happy enough @1395, just gonna let them roll while there r no issues and wait for the new addition to the family..
😉
 


If you have a driver crash, fault, or several memory errors the driver will downclock the card.
the only way to reset the driver and card is to reboot windows/Linux.

Running the Advanced/Beta flags increases the chance of this happening.

 


Well you tell me this as im testing my overclock right now @ 1508 Mhz (technically set to 1520 but it drops down to 1508). lol

max%20folding%20overclock%20core%2018_zpsvjisfs35.png~original
 

Seems the thermal throttle on the 980ti is around the 62-67C mark, mines @63-64C.

If you sit there and watch it warm up while folding you will see when it drops.

I did this and then reconfigured the fans using Afterburner to kick in just before this temp to keep the card running at the actual settings you apply.
 


Thanks for the info! didn't think it would drop the clocks slightly with temps limit being far from reach.

After watching my clocks and temps, Seams my card starts to drop around the 71-73C range
 

What is this thermal throttling of which you speak? My R9 290 has certainly never heard of such a thing. 😀

sh9.png

I should probably overclock at some point...
 


Yeah, the screenshot was after I had rebooted and my card went back to folding normally. The boost numbers I had posted came fro the specs for my card, which state "core clock 1216 MHz, boost clock 1367 MHz."

This is GTX 970 FTW, which I guess is OC'd from the factory. So maybe it won't tolerate any further OCing? It seems my card is running a bit higher by default than the stated factory settings. 1404MHz looks to be the norm -- HWMonitor says it has been as high as 1418.

 


You should keep an eye on different work units and see which ones push the clock speed up.

I'm surprised it goes as high as 1404 when the factory is 1367.

It should oc some more but if its running at 1404 and you added 20mhz to it it was attemptin to run @1424mhz which is quite a high clock setting, you may need a little tweak of extra voltage.

Keep an eye on wu - speed ratios and let us know, I'd leave it at defaults for now if its boosting that high.
 


overclocks that failed. At least i had hit the "Finish" option on each WU to make sure i didn't chew threw a ton of WU's.

So far 2 WU's was just fine on the 1510 MHz clock with a 10+ mV to assist in keeping things stable. the 56K WU and 120k WU

Although strangly, I could of sworn that 1 of the purposes of the F@Hbench program was to assist in overclocking as it simulated the same workload as a WU without using said WU's... (i know it's primarily for bench-marking but...)

It stop giving me errors at around 1545MHz.... Valley benchmark was more sensitive to around the 1535...

Oh well... At least things appear to be stable now.
 
Status
Not open for further replies.