F5F Stay Refreshed Power Users Overclocking 5.0GHz 7700k performance under high temperatures with HT enabled.

5.0GHz 7700k performance under high temperatures with HT enabled.

5.0GHz 7700k performance under high temperatures with HT enabled.

L
Leon1481
Member
53
06-12-2017, 08:17 PM
#1
I recently installed a brand new system (not much experienced with overclocking) and when testing Cinebench, my 7700k operates at 5.0GHz with a MAX package temperature of 77°C. The CPU-Z report displays only 1.25V on the CPU, assuming silent fan profiles in Q-FAN. This seems acceptable for a stress test. However, I can only reach this setting when HyperThreading is disabled. If I enable HT, the package temperature quickly rises to over 80°C before I stop the test! And that happens even with Turbo fan profiles active. Does this sound correct?

I suppose HT won’t be very helpful in everyday use, so I could still leave it off. For context, my 7700k is equipped with an NZXT Kraken X62 unit on the front panel and two 4-pin Corsair ML140 fans on the rear (push/pull). The rear fans connect to CHASSIS FAN 3 on my motherboard, while the front fans link to CHASSIS FAN 2.

In terms of overclocking steps, I’ve set the multiplier to 50 and the voltage to 1.370V.
L
Leon1481
06-12-2017, 08:17 PM #1

I recently installed a brand new system (not much experienced with overclocking) and when testing Cinebench, my 7700k operates at 5.0GHz with a MAX package temperature of 77°C. The CPU-Z report displays only 1.25V on the CPU, assuming silent fan profiles in Q-FAN. This seems acceptable for a stress test. However, I can only reach this setting when HyperThreading is disabled. If I enable HT, the package temperature quickly rises to over 80°C before I stop the test! And that happens even with Turbo fan profiles active. Does this sound correct?

I suppose HT won’t be very helpful in everyday use, so I could still leave it off. For context, my 7700k is equipped with an NZXT Kraken X62 unit on the front panel and two 4-pin Corsair ML140 fans on the rear (push/pull). The rear fans connect to CHASSIS FAN 3 on my motherboard, while the front fans link to CHASSIS FAN 2.

In terms of overclocking steps, I’ve set the multiplier to 50 and the voltage to 1.370V.

C
Catgirl546
Member
54
06-15-2017, 11:45 PM
#2
Hyperthreading leads to a slight rise in temperature. This is considered typical.
The value of HT varies depending on your activities.
Gaming has little impact.
Tasks like video encoding gain advantages.
Disable HT for gaming.
The fan positioning could actually reduce airflow if one side moves less air.
C
Catgirl546
06-15-2017, 11:45 PM #2

Hyperthreading leads to a slight rise in temperature. This is considered typical.
The value of HT varies depending on your activities.
Gaming has little impact.
Tasks like video encoding gain advantages.
Disable HT for gaming.
The fan positioning could actually reduce airflow if one side moves less air.

J
JR_GAMER07
Posting Freak
915
06-24-2017, 07:15 PM
#3
Hyperthreading leads to a slight rise in temperature. This is considered typical.
The value of HT varies based on your activities.
Gaming has little impact.
Tasks like video encoding gain advantages.
Disable HT for gaming.
The fan positioning could actually reduce airflow if one side moves less air.
J
JR_GAMER07
06-24-2017, 07:15 PM #3

Hyperthreading leads to a slight rise in temperature. This is considered typical.
The value of HT varies based on your activities.
Gaming has little impact.
Tasks like video encoding gain advantages.
Disable HT for gaming.
The fan positioning could actually reduce airflow if one side moves less air.

S
Shen907
Junior Member
9
07-10-2017, 08:24 AM
#4
Thanks, I believe I've actually made a small improvement. NZXT CAM was causing problems and not recognizing my Kraken cooler, but I think I resolved it so now I can set the Krakens pump speed to 'performance'.
I'm not convinced though—I'm seeing 5.0GHz during Cinebench with HT on, max package temp at 86°C, VCore 1.23 in CPU-Z (BIOS set to 1.37). A peak in the mid-80s is fine for a stress test on a 7700k, which should give mid to high 70s under normal conditions.
I ran Cinebench five times back-to-back without breaks, but maybe it's not the best stress test.
I'll probably turn off HT since my games run better without it and need a high clock speed for overclocking.
Maybe I'm missing something, but have I actually achieved what I was aiming for?
S
Shen907
07-10-2017, 08:24 AM #4

Thanks, I believe I've actually made a small improvement. NZXT CAM was causing problems and not recognizing my Kraken cooler, but I think I resolved it so now I can set the Krakens pump speed to 'performance'.
I'm not convinced though—I'm seeing 5.0GHz during Cinebench with HT on, max package temp at 86°C, VCore 1.23 in CPU-Z (BIOS set to 1.37). A peak in the mid-80s is fine for a stress test on a 7700k, which should give mid to high 70s under normal conditions.
I ran Cinebench five times back-to-back without breaks, but maybe it's not the best stress test.
I'll probably turn off HT since my games run better without it and need a high clock speed for overclocking.
Maybe I'm missing something, but have I actually achieved what I was aiming for?

D
deadly_cute
Member
59
07-10-2017, 11:42 AM
#5
And regarding the push/pull fan arrangement on my CPU radiator, you're right but I'll continue researching further. All four fans use 4-pin 140s and are set to operate under identical settings. The only uncertainties are if the two distinct models share the same airflow rate (unlikely) and whether varying the flow rate would actually impact performance. In the worst case, I might add another pair of ML140s so every fan on that radiator works together.
D
deadly_cute
07-10-2017, 11:42 AM #5

And regarding the push/pull fan arrangement on my CPU radiator, you're right but I'll continue researching further. All four fans use 4-pin 140s and are set to operate under identical settings. The only uncertainties are if the two distinct models share the same airflow rate (unlikely) and whether varying the flow rate would actually impact performance. In the worst case, I might add another pair of ML140s so every fan on that radiator works together.

1
14w1z0r
Junior Member
3
07-12-2017, 05:46 AM
#6
I would rather never exceed 80c even during a stress test. But if things are stable, skip the test and try something like x264 or real-world applications to check the temperatures.
If it works smoothly and stays cool, then yes.
Honestly, I can't provide a precise answer without knowing the fan's CFM rating. Just an observation.
1
14w1z0r
07-12-2017, 05:46 AM #6

I would rather never exceed 80c even during a stress test. But if things are stable, skip the test and try something like x264 or real-world applications to check the temperatures.
If it works smoothly and stays cool, then yes.
Honestly, I can't provide a precise answer without knowing the fan's CFM rating. Just an observation.

Z
Zack3823
Junior Member
2
08-01-2017, 10:47 AM
#7
I'd rather stay under 80c even during a stress test. But if it's stable and you get the desired speed without overheating, then go ahead.
If you don't know the fan power in CFM, I can't give a precise answer—just an observation.
Thanks for the tip, I'll try x264 next and maybe reduce the OC afterward.
Thanks!
Z
Zack3823
08-01-2017, 10:47 AM #7

I'd rather stay under 80c even during a stress test. But if it's stable and you get the desired speed without overheating, then go ahead.
If you don't know the fan power in CFM, I can't give a precise answer—just an observation.
Thanks for the tip, I'll try x264 next and maybe reduce the OC afterward.
Thanks!

C
Christy82
Member
52
08-01-2017, 02:03 PM
#8
I was wondering if dialing the OC back to 4.8GHz would help with x264. Even with the Krakens fans and full power at 100%, the package still reached 85°C maximum. The Vcore only climbed up to 1.270v during the test? I’m not sure why that is, since I’ve heard the Kraken should work well at 5.0GHz on the 7700k. I’m trying to figure out what I might be missing. Tomorrow I’ll put my 980ti in and see if gaming improves performance in a real-world test.
C
Christy82
08-01-2017, 02:03 PM #8

I was wondering if dialing the OC back to 4.8GHz would help with x264. Even with the Krakens fans and full power at 100%, the package still reached 85°C maximum. The Vcore only climbed up to 1.270v during the test? I’m not sure why that is, since I’ve heard the Kraken should work well at 5.0GHz on the 7700k. I’m trying to figure out what I might be missing. Tomorrow I’ll put my 980ti in and see if gaming improves performance in a real-world test.

M
MC_Asriel
Member
54
08-01-2017, 05:22 PM
#9
I understand many will differ, but if it works well in regular use then why worry about stress tests. Uncertain about the Vcore—perhaps it's limiting performance. If you have SpeedFan, it displays "Distance to tj max," indicating how close it is to throttling.
M
MC_Asriel
08-01-2017, 05:22 PM #9

I understand many will differ, but if it works well in regular use then why worry about stress tests. Uncertain about the Vcore—perhaps it's limiting performance. If you have SpeedFan, it displays "Distance to tj max," indicating how close it is to throttling.

D
dragon12600
Junior Member
10
08-01-2017, 07:27 PM
#10
I completely understand, it works well for my games without any crashes, that's what matters to me. I'll also check out Speedfan.
D
dragon12600
08-01-2017, 07:27 PM #10

I completely understand, it works well for my games without any crashes, that's what matters to me. I'll also check out Speedfan.