F5F Stay Refreshed Hardware Desktop Warning: AMD performance drops significantly when idle CPU usage rises.

Warning: AMD performance drops significantly when idle CPU usage rises.

Warning: AMD performance drops significantly when idle CPU usage rises.

J
JakeTVGaming
Senior Member
259
08-13-2016, 11:41 PM
#1
Reading a PDf or minimal activity? 30-40w. Full capacity? 90-120w. i5 2-310: running bf3 - 45w i5 idle: 5-10w. Is this correct? AMD power efficiency is good, but it can't drop below 3GHz.
J
JakeTVGaming
08-13-2016, 11:41 PM #1

Reading a PDf or minimal activity? 30-40w. Full capacity? 90-120w. i5 2-310: running bf3 - 45w i5 idle: 5-10w. Is this correct? AMD power efficiency is good, but it can't drop below 3GHz.

G
godzillaslime
Member
177
08-14-2016, 04:45 AM
#2
many programs and their mix can unexpectedly use a lot more CPU speed than needed. additionally, the way you measure power usage might vary, which could affect the results slightly.
G
godzillaslime
08-14-2016, 04:45 AM #2

many programs and their mix can unexpectedly use a lot more CPU speed than needed. additionally, the way you measure power usage might vary, which could affect the results slightly.

I
i3z___
Senior Member
559
08-24-2016, 08:17 PM
#3
Judy is utilizing HWM, and according to you, the change isn't significant.
I
i3z___
08-24-2016, 08:17 PM #3

Judy is utilizing HWM, and according to you, the change isn't significant.

C
chaospony
Member
149
08-24-2016, 09:32 PM
#4
It's not a valid comparison to directly pair these processors due to significant differences in architecture and capabilities.
C
chaospony
08-24-2016, 09:32 PM #4

It's not a valid comparison to directly pair these processors due to significant differences in architecture and capabilities.

M
MikeDragon159
Senior Member
661
08-25-2016, 09:39 PM
#5
Mine isn't much better. 12% of 540 watt = 65 watts for just the CPU. As you can see, all cores are asleep except for 2, one at 778 mhz and one at 274 mhz. Part of the reason is the large amount of RAM, at 3600 mhz. This I understand requires more frequent refreshes, and having 64 gigs means there is more memory to be refreshed. My previous skylake system when on an integrated GPU with 16 Gb of RAM would idle at 30-40 watts at the wall.
M
MikeDragon159
08-25-2016, 09:39 PM #5

Mine isn't much better. 12% of 540 watt = 65 watts for just the CPU. As you can see, all cores are asleep except for 2, one at 778 mhz and one at 274 mhz. Part of the reason is the large amount of RAM, at 3600 mhz. This I understand requires more frequent refreshes, and having 64 gigs means there is more memory to be refreshed. My previous skylake system when on an integrated GPU with 16 Gb of RAM would idle at 30-40 watts at the wall.

W
WaxOn_HaxOff
Junior Member
5
08-26-2016, 10:25 AM
#6
it runs quietly at 3 to 6 watts, showing about 102 watts total on the power meter. The desktop setup has a lot of RGB lighting in the background.
W
WaxOn_HaxOff
08-26-2016, 10:25 AM #6

it runs quietly at 3 to 6 watts, showing about 102 watts total on the power meter. The desktop setup has a lot of RGB lighting in the background.

N
NylodnewgPlaZ
Member
187
08-26-2016, 01:29 PM
#7
I believe the variation stems from how AMD and Intel handle reporting data. The user notes similarities between HWInfo and their own experience, where AMD's "CPU Package Power" reads about 30W, while their setup draws less than a comparable Intel system. Their system details—R9 5900X, RTX 2060 Super, cooling components, SSDs, and a 750W PSU—show idle power use between 82-89W as measured by UPS. The key distinctions are a less powerful graphics card, absence of an HDD, and no AIO pump, which likely explain the ~15-20W difference at idle.
N
NylodnewgPlaZ
08-26-2016, 01:29 PM #7

I believe the variation stems from how AMD and Intel handle reporting data. The user notes similarities between HWInfo and their own experience, where AMD's "CPU Package Power" reads about 30W, while their setup draws less than a comparable Intel system. Their system details—R9 5900X, RTX 2060 Super, cooling components, SSDs, and a 750W PSU—show idle power use between 82-89W as measured by UPS. The key distinctions are a less powerful graphics card, absence of an HDD, and no AIO pump, which likely explain the ~15-20W difference at idle.

X
xAuDesignsx
Member
214
08-27-2016, 02:29 AM
#8
ryzen 2600
X
xAuDesignsx
08-27-2016, 02:29 AM #8

ryzen 2600

B
Barrsy333
Junior Member
5
08-27-2016, 02:44 AM
#9
I highlighted the reported package power and then compared it to the actual system usage. For me, the 102w includes six RGB fans (3.6w each for the lights), an AIO pump, a RGB mouse and keyboard, a USB headset, and charging for USB earbuds. Most systems typically draw between 75 and 150w when idle.
B
Barrsy333
08-27-2016, 02:44 AM #9

I highlighted the reported package power and then compared it to the actual system usage. For me, the 102w includes six RGB fans (3.6w each for the lights), an AIO pump, a RGB mouse and keyboard, a USB headset, and charging for USB earbuds. Most systems typically draw between 75 and 150w when idle.

D
DarkTiger427
Member
186
08-27-2016, 03:54 AM
#10
The Z590 chipset states a TDP of 6 watts, but that figure can vary. Most components include heatsinks, so power use should be higher than the stated number. We estimate around 15 watts for safety. Adding keyboard, mouse, RGB fans, SSDs, and hard drives brings the total to roughly 80 watts. Including memory modules and USB charging, the overall load could reach 102 watts. At such low demand, the power supply efficiency drops, making the numbers seem accurate. Of course, these calculations are based on assumptions and may not reflect real performance.
D
DarkTiger427
08-27-2016, 03:54 AM #10

The Z590 chipset states a TDP of 6 watts, but that figure can vary. Most components include heatsinks, so power use should be higher than the stated number. We estimate around 15 watts for safety. Adding keyboard, mouse, RGB fans, SSDs, and hard drives brings the total to roughly 80 watts. Including memory modules and USB charging, the overall load could reach 102 watts. At such low demand, the power supply efficiency drops, making the numbers seem accurate. Of course, these calculations are based on assumptions and may not reflect real performance.