Question How to tell if overclock has actually been applied ?

karasahin

Distinguished
Sep 28, 2014
88
0
18,640
Hello,

I've increased my GPU's both core (from 600 to 750) and memory clocks (from 900 to 1050) and GPU-Z is reporting correct values. RivaTuner also is reporting the same in games etc.
So does this prove that the overclock has actually been applied ?

The reason I ask is that this overclock is only possible with a very old MSI Afterburner program since it's a mobile GPU (please don't lecture me about it, its temp doesn't go above 50C when supposedly overclocked).

Thanks.
 
Solution
99.9% not possible.

There are some differences when reporting temperatures. Sometimes differences between it and different monitoring programs, for example my memory speed was reported as 32 x 99.9. The motherboard frequency, bus speed, was detected as 99.9MHz so it didn’t report 3200M transfers but just a little under.

It’s seeing the clocks as you described. I’d say your overclock was active.
Download HWinfo, it monitors just about everything, provides logging and shows you the upper and lower bounds of each monitored part.

It’s free and lightweight, low processor overhead.

Run your programs and have a look.
Thanks for the suggestion. Also done that and the result is below. How possible is it that reporting values and actual values are not the same?
8INZlgD.png
 
99.9% not possible.

There are some differences when reporting temperatures. Sometimes differences between it and different monitoring programs, for example my memory speed was reported as 32 x 99.9. The motherboard frequency, bus speed, was detected as 99.9MHz so it didn’t report 3200M transfers but just a little under.

It’s seeing the clocks as you described. I’d say your overclock was active.
 
  • Like
Reactions: karasahin
Solution
99.9% not possible.

There are some differences when reporting temperatures. Sometimes differences between it and different monitoring programs, for example my memory speed was reported as 32 x 99.9. The motherboard frequency, bus speed, was detected as 99.9MHz so it didn’t report 3200M transfers but just a little under.

It’s seeing the clocks as you described. I’d say your overclock was active.
Thanks a lot.

One last thing: after overlocking, when you keep temps under control, experience no crashes in games/apps (no stability issues), will the GPU be as safe as if it's working in stock clocks? The GPU temps have always been superb so I'm lucky about that and had no issues with 750/1050 clocks for relatively long time.

750/1100 caused a crash to desktop in the last 3DMark 11 test and 780/1170 (maximum allowed for both) caused black pixels/spots in the same test so I took them as warnings and reverted back.

Maybe 775/1075 is the sweet spot but I haven't tried those.
 
So long as the temps are under control you should be good.
Extra heat will shorten the life of a chip. Thinking of old desktop CPUs the argument was that the lifetime of the cpu was, for example, a nominal 10 years and you intended to replace the system in 6 years then you may as well enjoy the extra performance as you wouldn’t degrade it by 4 years.

Make sure that the temps stay within a few degrees of the normal operating temperature and you should be ok. Also, run for a few days without the overclock.
Does it feel slow, can you tell the difference? If not then there is no benefit and don’t overclock.
 
  • Like
Reactions: karasahin
So long as the temps are under control you should be good.
Extra heat will shorten the life of a chip. Thinking of old desktop CPUs the argument was that the lifetime of the cpu was, for example, a nominal 10 years and you intended to replace the system in 6 years then you may as well enjoy the extra performance as you wouldn’t degrade it by 4 years.

Make sure that the temps stay within a few degrees of the normal operating temperature and you should be ok. Also, run for a few days without the overclock.
Does it feel slow, can you tell the difference? If not then there is no benefit and don’t overclock.
That's great to hear.

That's what I knew about extra heat as well, hence I used to turn off turbo boost to keep temps below 80C but when I brought this up in another place, I was told that this is not wise to do it and that extra heat won't shorten the life of the CPU (still a chip though right?) because of its design (throttles before reaching dangerous temp area).

So I turned turbo boost back on + overclocked the GPU but got a good cooling pad because I still want to keep temps below 80C for the CPU and 60C for the GPU. "So far so good" (summer is coming, goodbye to turbo boost and overclock soon probably).