F5F Stay Refreshed Power Users Overclocking the temperatures are a bit worrying.

the temperatures are a bit worrying.

the temperatures are a bit worrying.

S
Spoooky_
Junior Member
48
05-23-2020, 02:30 PM
#1
Hello everyone,
I’m currently operating at a 10600K and have been monitoring the temperatures closely. I’ll share as much detail as possible here.

My system is running a gentle 4.8ghz all-core overclock with a ring multiplier of 44x. I’m not using AVX offset, and the BIOS voltage is set at 1.280V, though with LLC settings, the actual Vcore under load drops to around 1.260V.

For cooling, I’m using a Noctua D15-S fan. The setup includes one fan for the front and two for the rear exhaust. I’m curious whether I’ve hit the thermal limits of this cooler—something that seems hard to believe—but I’m considering replacing it if needed. I’m leaning toward swapping it out, though I’d appreciate advice from other 10600K users who are also overclocking in air.

I’m particularly interested in how temperatures perform in applications like Prime95 26.6, Cinebench R20, Blender, etc. My motivation for this change comes from a past experience: during my initial build last week, I noticed higher-than-expected temps and realized something was amiss. It turned out the heat-sink mounting screw had only about three-quarters of a turn left before it stopped rising, and possibly there was an air gap between the cold plate and IHS. Tightening it further improved temperatures. However, this issue lasted for a couple of days before I realized it. I’m wondering if that slight unevenness affected my thermal paste application or if the temps are still due to cooling limitations.

Current readings:
- Ambient room: 22–23°C
- Idle: 35–38°C
- Prime95 26.6: 85–88°C
- Cinebench R20: 78–82°C
- AIDA64: 65–68°C
- Battlefield 5 Multiplayer (most CPU-heavy): 68–70°C
- Other single-threaded games: 50s range

My CPU’s P95 load power is 140W, and this cooler should handle well—likely around 230W minimum. Originally, I aimed for a 5.0GHz overclock at 1.35V or less, but now I’m settling at 4.8GHz due to limited headroom.

I believe this cooler is more than capable, but I want to confirm my setup. If anyone else is running a 10600K in air and shares their results, it would be helpful. I understand the temps under typical gaming use are safe, but I’m still curious about the thermal paste coverage and its impact on performance.

I plan to remove the cooler tomorrow and install fresh paste just to explore further. If you’re doing similar overclocking, please let me know your experiences and temperatures.
S
Spoooky_
05-23-2020, 02:30 PM #1

Hello everyone,
I’m currently operating at a 10600K and have been monitoring the temperatures closely. I’ll share as much detail as possible here.

My system is running a gentle 4.8ghz all-core overclock with a ring multiplier of 44x. I’m not using AVX offset, and the BIOS voltage is set at 1.280V, though with LLC settings, the actual Vcore under load drops to around 1.260V.

For cooling, I’m using a Noctua D15-S fan. The setup includes one fan for the front and two for the rear exhaust. I’m curious whether I’ve hit the thermal limits of this cooler—something that seems hard to believe—but I’m considering replacing it if needed. I’m leaning toward swapping it out, though I’d appreciate advice from other 10600K users who are also overclocking in air.

I’m particularly interested in how temperatures perform in applications like Prime95 26.6, Cinebench R20, Blender, etc. My motivation for this change comes from a past experience: during my initial build last week, I noticed higher-than-expected temps and realized something was amiss. It turned out the heat-sink mounting screw had only about three-quarters of a turn left before it stopped rising, and possibly there was an air gap between the cold plate and IHS. Tightening it further improved temperatures. However, this issue lasted for a couple of days before I realized it. I’m wondering if that slight unevenness affected my thermal paste application or if the temps are still due to cooling limitations.

Current readings:
- Ambient room: 22–23°C
- Idle: 35–38°C
- Prime95 26.6: 85–88°C
- Cinebench R20: 78–82°C
- AIDA64: 65–68°C
- Battlefield 5 Multiplayer (most CPU-heavy): 68–70°C
- Other single-threaded games: 50s range

My CPU’s P95 load power is 140W, and this cooler should handle well—likely around 230W minimum. Originally, I aimed for a 5.0GHz overclock at 1.35V or less, but now I’m settling at 4.8GHz due to limited headroom.

I believe this cooler is more than capable, but I want to confirm my setup. If anyone else is running a 10600K in air and shares their results, it would be helpful. I understand the temps under typical gaming use are safe, but I’m still curious about the thermal paste coverage and its impact on performance.

I plan to remove the cooler tomorrow and install fresh paste just to explore further. If you’re doing similar overclocking, please let me know your experiences and temperatures.

M
Malthe581
Member
140
05-23-2020, 06:12 PM
#2
You're on the correct track now. Adding an AIO will bring some noticeable improvement, and if you can adjust it slightly more, the results will be better. The 10600K offers a lot of flexibility that can be maximized even at just 1.3v. My 10700K reaches 5.1GHz at 1.295v, and with Turbo LLC and a 360mm AIO, performance remains solid on temperature management, indicating your configuration can be pushed further with proper cooling. An all-core 5GHz setup paired with a 280mm AIO should easily achieve temperatures that allow pushing the limits without relying heavily on AVX offsets.
M
Malthe581
05-23-2020, 06:12 PM #2

You're on the correct track now. Adding an AIO will bring some noticeable improvement, and if you can adjust it slightly more, the results will be better. The 10600K offers a lot of flexibility that can be maximized even at just 1.3v. My 10700K reaches 5.1GHz at 1.295v, and with Turbo LLC and a 360mm AIO, performance remains solid on temperature management, indicating your configuration can be pushed further with proper cooling. An all-core 5GHz setup paired with a 280mm AIO should easily achieve temperatures that allow pushing the limits without relying heavily on AVX offsets.

H
HoleInoneHusky
Junior Member
45
05-29-2020, 06:15 PM
#3
Remount the cooler... I'm not sure why you didn't shut everything down and fix it right away when you realized it wasn't fully secure. It could have moved easily from horizontal to vertical position since it wasn't properly locked in place.
2) No AVX offset, huh? You won't be able to achieve your 5.0ghz OC on that cooler if you don't.
At the same frequency as SSE, AVX is quicker but needs more voltage for stability—this also results in higher power consumption and increased heat generation.
A few applications—still a handful—run well, especially Cinebench R20...
H
HoleInoneHusky
05-29-2020, 06:15 PM #3

Remount the cooler... I'm not sure why you didn't shut everything down and fix it right away when you realized it wasn't fully secure. It could have moved easily from horizontal to vertical position since it wasn't properly locked in place.
2) No AVX offset, huh? You won't be able to achieve your 5.0ghz OC on that cooler if you don't.
At the same frequency as SSE, AVX is quicker but needs more voltage for stability—this also results in higher power consumption and increased heat generation.
A few applications—still a handful—run well, especially Cinebench R20...

D
dm20_tm
Member
227
05-31-2020, 04:18 AM
#4
I believe I would need a -1 or -2 AVX offset at 5.0ghz. I'm not currently running one at 4.8 across all cores. You're correct, I should have stopped it and remounted it right away when I noticed the issue with the mounting screws. It was late at night, and the temperatures improved, so I assumed everything was fine—though I should admit it was my mistake. Since then, I've started rethinking my decision-making process. I haven't had a chance to re-mount yet. I spent most of my free time today building another PC for a relative, but I'm still planning to do it tonight. I'll let you know once I've finished.
D
dm20_tm
05-31-2020, 04:18 AM #4

I believe I would need a -1 or -2 AVX offset at 5.0ghz. I'm not currently running one at 4.8 across all cores. You're correct, I should have stopped it and remounted it right away when I noticed the issue with the mounting screws. It was late at night, and the temperatures improved, so I assumed everything was fine—though I should admit it was my mistake. Since then, I've started rethinking my decision-making process. I haven't had a chance to re-mount yet. I spent most of my free time today building another PC for a relative, but I'm still planning to do it tonight. I'll let you know once I've finished.

E
Emmy149
Member
56
06-01-2020, 06:11 PM
#5
All good then. Keep us posted.
E
Emmy149
06-01-2020, 06:11 PM #5

All good then. Keep us posted.

C
Ciera7
Member
225
06-01-2020, 06:58 PM
#6
The re-mount has once more enhanced the temperature readings. Initially, I applied the Noctua NT-H1 paste provided with the cooler. This time, I opted for a brand new AS5 paste that I had purchased earlier when assembling this setup. I applied roughly one grain of rice in a straight line across the IHS. It's worth noting that I only switched to the AS5 just to experiment with a different paste, not because I was dissatisfied with the Noctua NT-H1 already installed on the CPU. I don't claim either paste is superior.

Idle and load temperatures have both seen improvement. Previously, idle temps hovered between 35-38°C, now they sit around 32-34°C. Load temperatures in Cinebench R20 and Prime95 26.6 show an improvement of about 4 to 6°C. I no longer reach over 80°C during R20 runs; the temps remain stable between 74-76°C, with a peak of 78°C on one core. For Prime95 26.6 Small FFT, the peak temperature on a single core is now 82°C, while all other cores stay near 78-80°C. These values are much more comfortable. This assessment was made at a room temperature of 23°C, and I was simply running standard fan curves for both CPU and case fans without pushing them to their limits beforehand.

It's difficult to judge the thermal paste quality after removing the cooler, as it seems I might have applied a bit too much initially. My assumption was that the lower part of the IHS could use better coverage, which aligned with my expectations given the uneven application. Still, seeing these improvements is encouraging.
C
Ciera7
06-01-2020, 06:58 PM #6

The re-mount has once more enhanced the temperature readings. Initially, I applied the Noctua NT-H1 paste provided with the cooler. This time, I opted for a brand new AS5 paste that I had purchased earlier when assembling this setup. I applied roughly one grain of rice in a straight line across the IHS. It's worth noting that I only switched to the AS5 just to experiment with a different paste, not because I was dissatisfied with the Noctua NT-H1 already installed on the CPU. I don't claim either paste is superior.

Idle and load temperatures have both seen improvement. Previously, idle temps hovered between 35-38°C, now they sit around 32-34°C. Load temperatures in Cinebench R20 and Prime95 26.6 show an improvement of about 4 to 6°C. I no longer reach over 80°C during R20 runs; the temps remain stable between 74-76°C, with a peak of 78°C on one core. For Prime95 26.6 Small FFT, the peak temperature on a single core is now 82°C, while all other cores stay near 78-80°C. These values are much more comfortable. This assessment was made at a room temperature of 23°C, and I was simply running standard fan curves for both CPU and case fans without pushing them to their limits beforehand.

It's difficult to judge the thermal paste quality after removing the cooler, as it seems I might have applied a bit too much initially. My assumption was that the lower part of the IHS could use better coverage, which aligned with my expectations given the uneven application. Still, seeing these improvements is encouraging.

R
ReborntoKill
Posting Freak
821
06-02-2020, 02:09 PM
#7
The top-tier pastes differ by about 1C – maybe that’s the key.
Applying the label ‘high end’ to thermal paste seems a bit odd.
It might just be that adjusting the mount was the real solution.
R
ReborntoKill
06-02-2020, 02:09 PM #7

The top-tier pastes differ by about 1C – maybe that’s the key.
Applying the label ‘high end’ to thermal paste seems a bit odd.
It might just be that adjusting the mount was the real solution.

P
Pochisa
Junior Member
43
06-07-2020, 03:29 AM
#8
You're on the correct track now. Adding an AIO will bring some noticeable improvement, and if you can adjust it slightly more, the results will be better. The 10600K offers a lot of flexibility that can be maximized even at just 1.3v. My 10700K reaches 5.1GHz at 1.295v, and with Turbo LLC and a 360mm AIO, performance remains solid on temperature management—demonstrating that with proper cooling, your configuration can be pushed further. An all-core 5GHz setup paired with a 280mm AIO should easily achieve temperatures well within control, even without AVX adjustments.
P
Pochisa
06-07-2020, 03:29 AM #8

You're on the correct track now. Adding an AIO will bring some noticeable improvement, and if you can adjust it slightly more, the results will be better. The 10600K offers a lot of flexibility that can be maximized even at just 1.3v. My 10700K reaches 5.1GHz at 1.295v, and with Turbo LLC and a 360mm AIO, performance remains solid on temperature management—demonstrating that with proper cooling, your configuration can be pushed further. An all-core 5GHz setup paired with a 280mm AIO should easily achieve temperatures well within control, even without AVX adjustments.