F5F Stay Refreshed Power Users Overclocking High v-droop in the maximus xi equation with LLC7

High v-droop in the maximus xi equation with LLC7

High v-droop in the maximus xi equation with LLC7

S
Seby777
Member
Find
58
12-04-2018, 03:22 AM
#1
Hi
I've been testing my new Asus Maximus XI formula and 9900k with 5GHz all-core, 1.33V BIOS and LLc7. It still drops to 1.279V under load. Are there any settings I overlooked? Should I be using the maximum LLc8? Did I receive a faulty board?
Reply
Reply
S
Seby777
12-04-2018, 03:22 AM #1

Hi
I've been testing my new Asus Maximus XI formula and 9900k with 5GHz all-core, 1.33V BIOS and LLc7. It still drops to 1.279V under load. Are there any settings I overlooked? Should I be using the maximum LLc8? Did I receive a faulty board?

Reply
Reply
T
TorcikPL
Member
Find
177
12-21-2018, 04:16 AM
#2
Just to be sure, you're sure its Vdroop? maybe its c-state playing around? on load you mean tests such as prime? if so, I think thats a pretty huge variation and LLC ain't keeping up, you can try it at max just to see what happens, the real problem would be if it were applying too much compensation volts, wich doesnt seem to be the case.
Reply
Reply
T
TorcikPL
12-21-2018, 04:16 AM #2

Just to be sure, you're sure its Vdroop? maybe its c-state playing around? on load you mean tests such as prime? if so, I think thats a pretty huge variation and LLC ain't keeping up, you can try it at max just to see what happens, the real problem would be if it were applying too much compensation volts, wich doesnt seem to be the case.

Reply
Reply
D
DX_GamerTV
Junior Member
Find
7
12-21-2018, 08:02 AM
#3
Rodrigodrt :
Just to confirm, are you certain it's Vdroop? Could it be some c-state behavior? On load you're referring to tests like prime? If that's the case, it seems like a significant variation, and LLC isn't handling it well. You might want to try running it at its maximum just to observe what happens. The real issue would likely be excessive compensation voltages, which doesn't appear to be the problem here.

I think so. Using prime95 with non-avx small FFT to stress it works too—similar results in other programs as well. I eventually tried LLC8 but it didn't help. It caused a BSOD.
Reply
Reply
D
DX_GamerTV
12-21-2018, 08:02 AM #3

Rodrigodrt :
Just to confirm, are you certain it's Vdroop? Could it be some c-state behavior? On load you're referring to tests like prime? If that's the case, it seems like a significant variation, and LLC isn't handling it well. You might want to try running it at its maximum just to observe what happens. The real issue would likely be excessive compensation voltages, which doesn't appear to be the problem here.

I think so. Using prime95 with non-avx small FFT to stress it works too—similar results in other programs as well. I eventually tried LLC8 but it didn't help. It caused a BSOD.

Reply
Reply