I've got a 960 SSC and when the GPU core hits 60c the voltage throttles from 1253mv to 1237mv and the core overclock drops as well. This is happening despite the fact that the temp target is set to 95c. Is there any reason this is happening and is there any way around it?

rjohnson11
Can you retest this with your CPU at default clocks?

Still throttled, in fact this time it happened at 59c instead of 60c.

mapoetti
I have the similar problem.     what is?   please help

I've been speaking to a lot of people about this and it appears to be a "feature" rather than an issue in Maxwell chips. Several people I've talked to spoke of the voltage in the 970's throttling at/around 68c. I honestly can't for the life of me figure out why it would be set so low on the 960. Because of this it means I have to run the fans on the card anywhere between 60-70% to keep the temps below 60c so I don't lose performance and this is awful because the ACX cooler is disgustingly loud. 
 
Since the 960 I have was obtained through an unwanted upgrade via RMA I have only a few options around this other than than selling the 960 for something. I was running 760 SLI in my system and my other card is in the process of being replaced by a 960 as well and this will only make the noise issues worse as there will be two cards that have to run with cranked fans instead of just one. This will make my system unacceptably loud compared to how it was before with my 760's as I could run them with a max fan speed of 45% and not have to worry about throttling because the 760's don't throttle at a pants-on-head retarded temperature.
 
I already wasn't going to buy another EVGA product after the ridiculous RMA fiasco I've gone through but apparently EVGA thought that I needed even more reasons to avoid their hardware in the future.

mapoetti

se7enupmustang

mapoetti
I have the similar problem.     what is?   please help

I've been speaking to a lot of people about this and it appears to be a "feature" rather than an issue in Maxwell chips. Several people I've talked to spoke of the voltage in the 970's throttling at/around 68c. I honestly can't for the life of me figure out why it would be set so low on the 960. Because of this it means I have to run the fans on the card anywhere between 60-70% to keep the temps below 60c so I don't lose performance and this is awful because the ACX cooler is disgustingly loud. 
 
Since the 960 I have was obtained through an unwanted upgrade via RMA I have only a few options around this other than than selling the 960 for something. I was running 760 SLI in my system and my other card is in the process of being replaced by a 960 as well and this will only make the noise issues worse as there will be two cards that have to run with cranked fans instead of just one. This will make my system unacceptably loud compared to how it was before with my 760's as I could run them with a max fan speed of 45% and not have to worry about throttling because the 760's don't throttle at a pants-on-head retarded temperature.
 
I already wasn't going to buy another EVGA product after the ridiculous RMA fiasco I've gone through but apparently EVGA thought that I needed even more reasons to avoid their hardware in the future.

ok i understend.
in this moment i have do a test.  i have switch  the bios   and  i have  a bad news:  in  normal  mode   video card run ok. In performance  base clock  default not change. in to afterburner program  i see power limit 110 max  and no 115%.
 in normal mode all ok  i  see 115% power limit
 
problem is  in performance mode   bios switch ,power limit worck but  frequency stay at default....o_0
i remember  first day   the bios switch worck good.
 
your bios switch worck?

My switch works, in the first bios I have a 110% power limit and a 115% in the second bios but my voltage throttle point stays the same.

Shelledfade
this is that boost power savings crap.

I get the same thing on my gtx 980 at 67-68c where it will throttle voltage down from 1.2180 to 1.1930 

I've bypassed this annoyance by adding +20mV voltage with afterburner, so that when this happens I'll still be at my highest defaulted voltage.

Doesn't help me though as im getting throttled from my max oc voltage at 60c which is very much what I dont bloody want to happen.

Try changing the power management mode in GeForce control panel to "Prefer max performance"

Let me get this correct, you're running +90 on the core and +675 on the memory?...Most likely you are exceeding the power limit. Your 'access denied' is likely a permissions issue with your OS. After all, 99.99% of people using the drivers don't have that issue. Have you tried doing a clean driver install using DDU?
http://forums.guru3d.com/showthread.php?t=379506

To fix your access is denied error check the security premissions for folders "nvidia corporation" and "drs" under c:\programdata\nvidia corporation\drs. All users & groups should have full control check marked.

se7enupmustang
I've got a 960 SSC and when the GPU core hits 60c the voltage throttles from 1253mv to 1237mv and the core overclock drops as well. This is happening despite the fact that the temp target is set to 95c. Is there any reason this is happening and is there any way around it?

The behavior you are seeing is perfectly normal for GPU Boost 2.0.

http://forums.evga.com/98...bove-60C-m2355251.aspx

Maxwell cards have GPU Boost 2.0 which boosts substantially higher than GPU Boost 1.0, but the boost curve incrementally falls at a slightly steeper rate. Such tiny drops in core voltage and boost clock should be no concern of yours.



Throttling doesn't occur until max temperature is reached. "Throttling" is when the core clock drops below base clock speed. If there were another bar to the right in the graph above, that next bar would show throttling when the core temperature continued to increase. Incremental drops in boost clock is NOT throttling.

se7enupmustang

ty_ger07

se7enupmustang
I've got a 960 SSC and when the GPU core hits 60c the voltage throttles from 1253mv to 1237mv and the core overclock drops as well. This is happening despite the fact that the temp target is set to 95c. Is there any reason this is happening and is there any way around it?

The behavior you are seeing is perfectly normal for GPU Boost 2.0.

http://forums.evga.com/98...bove-60C-m2355251.aspx

Maxwell cards have GPU Boost 2.0 which boosts substantially higher than GPU Boost 1.0, but the boost curve incrementally falls at a slightly steeper rate. Such tiny drops in core voltage and boost clock should be no concern of yours.



Throttling doesn't occur until max temperature is reached. "Throttling" is when the core clock drops below base clock speed. If there were another bar to the right in the graph above, that next bar would show throttling when the core temperature continued to increase. Incremental drops in boost clock is NOT throttling.

Considering that drops from my desired clock at 60c I'm calling it throttling. There's no reason for it to drop at that temperature, It's already slower than the card it replaced so it getting slower after that is a concern. How is that hard to understand?

Call it whatever you want. You would be calling it incorrectly, but it is your right.

What am I not understanding? What are YOU not understanding? I understand your concern, but I am telling you that is just the way it is and there is little you can do but live with it. It is working the way it was designed.

Good day!