F5F Stay Refreshed Hardware Desktop Performance stays at 5.25 GHz or lower

Performance stays at 5.25 GHz or lower

Performance stays at 5.25 GHz or lower

Pages (2): 1 2 Next
Z
zlip88
Member
198
11-20-2022, 01:13 AM
#1
Hello everyone, here’s a quick summary: I’m used to older hardware and just upgraded a couple of months ago. My current setup is an AMD Ryzen 9 7900X with a boost clock around 5.6 GHz. Despite heavy use—gaming, work, or simulations—I’ve never reached speeds past 5.25 GHz. The temperatures stay manageable at about 80°C under load. Everything in the UEFI is set to defaults except EXPO, and I’ve turned off the built-in GPU. My drivers and firmware are up to date. I’m wondering if there’s something I’m missing to fully utilize this processor. Any tips would be appreciated!
Z
zlip88
11-20-2022, 01:13 AM #1

Hello everyone, here’s a quick summary: I’m used to older hardware and just upgraded a couple of months ago. My current setup is an AMD Ryzen 9 7900X with a boost clock around 5.6 GHz. Despite heavy use—gaming, work, or simulations—I’ve never reached speeds past 5.25 GHz. The temperatures stay manageable at about 80°C under load. Everything in the UEFI is set to defaults except EXPO, and I’ve turned off the built-in GPU. My drivers and firmware are up to date. I’m wondering if there’s something I’m missing to fully utilize this processor. Any tips would be appreciated!

X
xXRAXERXx
Posting Freak
817
11-20-2022, 08:53 AM
#2
The maximum speed increase applies to one core only. When multiple cores are active, performance tends to drop. Have you checked if any of the synthetic tasks ran in a single-threaded mode, such as Cinebench Single Core?
X
xXRAXERXx
11-20-2022, 08:53 AM #2

The maximum speed increase applies to one core only. When multiple cores are active, performance tends to drop. Have you checked if any of the synthetic tasks ran in a single-threaded mode, such as Cinebench Single Core?

P
Plant_b
Member
50
11-20-2022, 10:04 AM
#3
I also performed a Cinebench (2024) single-core benchmark right before beginning this discussion, which returned a score of 117—uncertain if typical for this processor. The test consistently stayed below 5.25 GHz.
P
Plant_b
11-20-2022, 10:04 AM #3

I also performed a Cinebench (2024) single-core benchmark right before beginning this discussion, which returned a score of 117—uncertain if typical for this processor. The test consistently stayed below 5.25 GHz.

T
TrixyElli
Member
52
11-24-2022, 12:27 PM
#4
In general terms, the numbers are reasonable. Hardware Unboxed reported a 124 for their 7900X compared to the 9950X, which is just a slight improvement. You might want to turn on PBO, but it doesn’t appear you’re running low on power, so any effect should be small. It’s unclear what else to test at this point. It looks like you’ve addressed most of the topics in your earlier review.
T
TrixyElli
11-24-2022, 12:27 PM #4

In general terms, the numbers are reasonable. Hardware Unboxed reported a 124 for their 7900X compared to the 9950X, which is just a slight improvement. You might want to turn on PBO, but it doesn’t appear you’re running low on power, so any effect should be small. It’s unclear what else to test at this point. It looks like you’ve addressed most of the topics in your earlier review.

B
BrickArms22123
Member
72
12-13-2022, 02:46 PM
#5
Yes, you can manually set the ratio for all cores.
B
BrickArms22123
12-13-2022, 02:46 PM #5

Yes, you can manually set the ratio for all cores.

T
TheKawaiiMelon
Junior Member
46
12-30-2022, 10:14 AM
#6
It’s crucial to know whether you performed a new setup or carried over your operating system from the old machine. When moving between systems, especially those with different processors, a clean installation is recommended.
T
TheKawaiiMelon
12-30-2022, 10:14 AM #6

It’s crucial to know whether you performed a new setup or carried over your operating system from the old machine. When moving between systems, especially those with different processors, a clean installation is recommended.

T
Tijeyd
Member
189
01-07-2023, 02:45 PM
#7
Okay, I tried setting PBO to the "Enhancement" setting on Level 1 (is that the correct way to go about it?). So far no change in results though Yes, it's a completely fresh Windows 11 installation. Just for reference, here's a screenshot I took during a Cinebench run after enabling PBO (I quickly took it when I saw the 5.25 GHz coming along): The test finished with a score of 118
T
Tijeyd
01-07-2023, 02:45 PM #7

Okay, I tried setting PBO to the "Enhancement" setting on Level 1 (is that the correct way to go about it?). So far no change in results though Yes, it's a completely fresh Windows 11 installation. Just for reference, here's a screenshot I took during a Cinebench run after enabling PBO (I quickly took it when I saw the 5.25 GHz coming along): The test finished with a score of 118

H
hdoor20
Senior Member
477
01-07-2023, 03:55 PM
#8
Have you considered using HWInfo64 for monitoring? It records peaks and troughs, giving a clearer view of the actual maximum frequency, and it performs better than Task Manager.
H
hdoor20
01-07-2023, 03:55 PM #8

Have you considered using HWInfo64 for monitoring? It records peaks and troughs, giving a clearer view of the actual maximum frequency, and it performs better than Task Manager.

S
Say_Say123
Member
120
01-09-2023, 03:44 AM
#9
It seems you're questioning whether you've been misled by Windows. Are you considering reverting PBO to automatic now that you're seeing improvements?
S
Say_Say123
01-09-2023, 03:44 AM #9

It seems you're questioning whether you've been misled by Windows. Are you considering reverting PBO to automatic now that you're seeing improvements?

A
Amtrak10
Senior Member
639
01-10-2023, 02:36 PM
#10
You can turn off PBO. It only becomes relevant in power-limited situations, which shouldn't apply to a single-core load. Those figures seem more aligned with what I'd expect from a 7900X. Task Manager might have been underreporting and led you on a wild goose chase.
A
Amtrak10
01-10-2023, 02:36 PM #10

You can turn off PBO. It only becomes relevant in power-limited situations, which shouldn't apply to a single-core load. Those figures seem more aligned with what I'd expect from a 7900X. Task Manager might have been underreporting and led you on a wild goose chase.

Pages (2): 1 2 Next