F5F Stay Refreshed Hardware Desktop Did the overclocking damage your CPU?

Did the overclocking damage your CPU?

Did the overclocking damage your CPU?

Pages (3): Previous 1 2 3
B
Bayan9
Member
158
05-19-2019, 01:37 PM
#21
What voltage are you using now? What cooler are you running? Prime95 small ffts is a real test to ensure your CPU doesn’t overheat, so I recommend AIDA64 with P95 large FFTs. Personally, I stick with P95 large FFTs alone, assuming the CPU voltage should match the frequency. If it’s unstable, you can slightly increase it. For stable operation, use the average voltage for that frequency on your specific CPU model—it usually works well. You can boost it a bit if needed, but keep it under 75°C to avoid wasted cooling. Run them as fast as possible without making noise; otherwise, you’re missing efficiency gains.
B
Bayan9
05-19-2019, 01:37 PM #21

What voltage are you using now? What cooler are you running? Prime95 small ffts is a real test to ensure your CPU doesn’t overheat, so I recommend AIDA64 with P95 large FFTs. Personally, I stick with P95 large FFTs alone, assuming the CPU voltage should match the frequency. If it’s unstable, you can slightly increase it. For stable operation, use the average voltage for that frequency on your specific CPU model—it usually works well. You can boost it a bit if needed, but keep it under 75°C to avoid wasted cooling. Run them as fast as possible without making noise; otherwise, you’re missing efficiency gains.

F
FELIPE369
Member
234
05-21-2019, 10:47 AM
#22
I mentioned lowering the CPU voltage until it becomes unstable at normal speed. Then gradually raise the multiplier and V settings as required. Certain processors handle overclocking well even without a voltage boost, so it’s worth giving it a shot if you’re willing to invest some time. You run Prime 95 and select small FFTs for testing. Execute it for 15 minutes; if any issues arise (P95 crashes, error messages, blue screen, or freezing), it indicates instability. EDIT: Regarding temperatures, you shouldn’t notice readings above 100°C since the CPU should automatically throttle. Aim to keep below 80°C if possible, but under 90°C is acceptable though it might reduce lifespan. If you’re seeing 100°C at stock speed, consider applying thermal paste or re-seating your cooler—it shouldn’t get that hot. Generally, the latest model is just a minor overclock of the previous version with slight benefits as noted by Somerandomtechyboi. Concerning small versus large FFTs, I’m accustomed to using small ones to push the CPU to its limits and force errors out. If no issues appear, everything should perform well regardless of circumstances. For more realistic testing, Cinebench is a free alternative, though I personally prefer it mainly for benchmarking rather than stability checks. He has the Noctua NH-U12S model.
F
FELIPE369
05-21-2019, 10:47 AM #22

I mentioned lowering the CPU voltage until it becomes unstable at normal speed. Then gradually raise the multiplier and V settings as required. Certain processors handle overclocking well even without a voltage boost, so it’s worth giving it a shot if you’re willing to invest some time. You run Prime 95 and select small FFTs for testing. Execute it for 15 minutes; if any issues arise (P95 crashes, error messages, blue screen, or freezing), it indicates instability. EDIT: Regarding temperatures, you shouldn’t notice readings above 100°C since the CPU should automatically throttle. Aim to keep below 80°C if possible, but under 90°C is acceptable though it might reduce lifespan. If you’re seeing 100°C at stock speed, consider applying thermal paste or re-seating your cooler—it shouldn’t get that hot. Generally, the latest model is just a minor overclock of the previous version with slight benefits as noted by Somerandomtechyboi. Concerning small versus large FFTs, I’m accustomed to using small ones to push the CPU to its limits and force errors out. If no issues appear, everything should perform well regardless of circumstances. For more realistic testing, Cinebench is a free alternative, though I personally prefer it mainly for benchmarking rather than stability checks. He has the Noctua NH-U12S model.

C
CrEaTuRe1312
Member
51
05-23-2019, 04:45 AM
#23
Yeah, I also saw that section before, but somehow missed it for a few seconds. It sounds like a burn test boosted to high power, making the 100c look pretty obvious.
C
CrEaTuRe1312
05-23-2019, 04:45 AM #23

Yeah, I also saw that section before, but somehow missed it for a few seconds. It sounds like a burn test boosted to high power, making the 100c look pretty obvious.

C
crazyborg
Member
122
05-27-2019, 09:23 PM
#24
Sure, I'll make sure to follow your instructions and aim for the target speed! Thanks a lot!
C
crazyborg
05-27-2019, 09:23 PM #24

Sure, I'll make sure to follow your instructions and aim for the target speed! Thanks a lot!

M
molresa
Junior Member
26
05-27-2019, 09:38 PM
#25
You're only observing 82°C after five minutes of small FFTs, not the smallest ones. It seems the signal was rising gradually from 80°, so a 100° reading shouldn't appear. Thanks! Good luck.
M
molresa
05-27-2019, 09:38 PM #25

You're only observing 82°C after five minutes of small FFTs, not the smallest ones. It seems the signal was rising gradually from 80°, so a 100° reading shouldn't appear. Thanks! Good luck.

Pages (3): Previous 1 2 3