F5F Stay Refreshed Hardware Desktop The CPU idle power consumption varies based on the motherboard.

The CPU idle power consumption varies based on the motherboard.

The CPU idle power consumption varies based on the motherboard.

R
RoboTron56
Junior Member
37
03-02-2018, 06:47 PM
#1
I have an Intel i7 8700 CPU without a dedicated memory controller. I've tried it on two different motherboards. The Asus Prime Z390-A board uses the Celeron processor, which can be replaced with the i7 8700. It seems the idle power consumption reported by tools like LibreHardwareMonitor and HWmonitor is significantly higher in the Asus board—around 10 to 14 watts—compared to just 3.5 to 4 watts on the HP Slimline 290-p0043w system. I’m puzzled about this difference. Why does the CPU draw more power at idle on the Asus board than it does on the HP model? This test was performed without any graphics cards, using only the built-in graphics and a single 1080p display while sitting in a Windows desktop idle session. Only one mouse and keyboard were connected via USB, and the "minimum processor state" in Windows was set to 0% on both systems.
R
RoboTron56
03-02-2018, 06:47 PM #1

I have an Intel i7 8700 CPU without a dedicated memory controller. I've tried it on two different motherboards. The Asus Prime Z390-A board uses the Celeron processor, which can be replaced with the i7 8700. It seems the idle power consumption reported by tools like LibreHardwareMonitor and HWmonitor is significantly higher in the Asus board—around 10 to 14 watts—compared to just 3.5 to 4 watts on the HP Slimline 290-p0043w system. I’m puzzled about this difference. Why does the CPU draw more power at idle on the Asus board than it does on the HP model? This test was performed without any graphics cards, using only the built-in graphics and a single 1080p display while sitting in a Windows desktop idle session. Only one mouse and keyboard were connected via USB, and the "minimum processor state" in Windows was set to 0% on both systems.

M
MrSuslik
Member
58
03-02-2018, 08:03 PM
#2
Various voltage charts, compatibility with multiple C states, diverse chip designs (still undecided about the board), improved power regulators, additional internal parts. A wide range of features overall.
M
MrSuslik
03-02-2018, 08:03 PM #2

Various voltage charts, compatibility with multiple C states, diverse chip designs (still undecided about the board), improved power regulators, additional internal parts. A wide range of features overall.

C
Cam_
Junior Member
18
03-03-2018, 12:12 AM
#3
Alternative C state management at the UEFI layer could change how the system behaves. HP might restrict CPU performance and favor a specific variant like the 8700 over a custom-built motherboard. I suspect the scores for single-core and all-core benchmarks would differ noticeably even with identical heatsinks. Whether each machine used the same boot drive or OS is uncertain.
C
Cam_
03-03-2018, 12:12 AM #3

Alternative C state management at the UEFI layer could change how the system behaves. HP might restrict CPU performance and favor a specific variant like the 8700 over a custom-built motherboard. I suspect the scores for single-core and all-core benchmarks would differ noticeably even with identical heatsinks. Whether each machine used the same boot drive or OS is uncertain.

T
Twothemoon
Junior Member
26
03-03-2018, 01:34 AM
#4
They had various boot OS drives, but I rely on scripts to configure them consistently. There could be other factors contributing to the issue, so I might re-test using the same drive later. It doesn’t appear the HP restricts the 8700 model. The device earned a multi Cinebench score of 7500 mainly due to early thermal throttling from the inadequate default cooler included.
T
Twothemoon
03-03-2018, 01:34 AM #4

They had various boot OS drives, but I rely on scripts to configure them consistently. There could be other factors contributing to the issue, so I might re-test using the same drive later. It doesn’t appear the HP restricts the 8700 model. The device earned a multi Cinebench score of 7500 mainly due to early thermal throttling from the inadequate default cooler included.

H
HydraEliphant
Member
212
03-07-2018, 12:47 PM
#5
Are you installing HP's original equipment manufacturer Windows for the slimline model? This setup might have a different scheduler setup that could limit certain performance modes, making the i7 8700 behave more like an i7 8700T. You may be able to improve cooling by adding heatsinks or mounting the CPU in a way it fits properly. Alternatively, taking the machine apart and working on it from a workbench could also help.
H
HydraEliphant
03-07-2018, 12:47 PM #5

Are you installing HP's original equipment manufacturer Windows for the slimline model? This setup might have a different scheduler setup that could limit certain performance modes, making the i7 8700 behave more like an i7 8700T. You may be able to improve cooling by adding heatsinks or mounting the CPU in a way it fits properly. Alternatively, taking the machine apart and working on it from a workbench could also help.

B
brobear7
Posting Freak
892
03-07-2018, 01:58 PM
#6
It wasn't the standard Windows setup that came with the device because Linux was already installed during purchase on Ebay. I added Windows myself, which likely explains the strong Cinebench result (the 8700T might score lower).
B
brobear7
03-07-2018, 01:58 PM #6

It wasn't the standard Windows setup that came with the device because Linux was already installed during purchase on Ebay. I added Windows myself, which likely explains the strong Cinebench result (the 8700T might score lower).