2080 Ti - power TDP drop in Firestrike.

[and23]

Hello. My card on stock,no oc. Aorus Rtx 2080 Ti Wateforce Xtreme. Today i run firestrike and looking on msi afterburner. For a moment ( exactly 1 second ) gpu power drop for 1 second and clock ,look.

Here,gpu power 58% and clock 1575mhz,but gpu load 96% ( why 96% ? ):

https://i.postimg.cc/KYpd9ZJv/gfgf.jpg

And bum normal again:

Normal clock, 1980mhz:



https://i.postimg.cc/Dfr0WF0r/ui2.jpg


Its not psu or gpu issue? I ask because like you see on diagram.


1 screen.
Red arrows. First time gpu power drop to 20%,GPU LOAD stays 96% ,but clock steady 1980mhz
Red arrows 2: Gpu power drop to 58%,GPU LOAD still stays 96%, but this time clock dropped to 1575 mhz for a second.

And why then GPU LOAD not dropped?

Anyway i cant reproduce this again with the same apps in background. Now its smooth without that stutter,weir.

But its was not hardware related?




PC:
Asus Prime Z390-A ( bios updated )
2x16 GB DDR4 GSKILL 3000MHZ XMP
Corsair 850RmX
9900K 4.7 ghz stock normal temps
Rtx 2080 Ti Waterforce Xtreme ,stock
SSD Crucial 1TB
HDD 1TB 7200rpm
Drivers was 436.15

Windows 10 1903v

[-Scud-]

Die Amtssprache ist deutsch!

[Maorga]

Its not psu or gpu issue? I ask because like you see on diagram.


1 screen.
Red arrows. First time gpu power drop to 20%,GPU LOAD stays 96% ,but clock steady 1980mhz
Red arrows 2: Gpu power drop to 58%,GPU LOAD still stays 96%, but this time clock dropped to 1575 mhz for a second.

And why then GPU LOAD not dropped?

But its was not hardware related?




It is definitely not a PSU problem. If the PSU could not handle the power consumption it would shut down. So you PC would be off.
Why in the second position the GPU drops I can't say, could be a stutter like loading textures, cpu is to late or something else. Could also be a driver Issue. But it is not Hardware related.

Regards

Norbert

[and23]

I mean why TDP ( power ) dropped from 100% to 28%. And second time to 58%.
But gpu load was always 96%.

[Maorga]

Maybe the polling is to high reading the temperature. It could be caused by many things, also by the cpu.
But your PSU nor GPU is defective.

[and23]

Maybe the polling is to high reading the temperature. It could be caused by many things, also by the cpu.
But your PSU nor GPU is defective.
So it was not caused by faulty psu or gpu?

[Maorga]

No definitely not

[and23]

Hey guys. Just do not want to refresh all topic. I tested time spy gpu test 1.

I run TimeSpy looped gpu test 1. Clock during 40 minutes was 1920-1970mhz with random power usage. But one time it dropped to 1800mhz. Is this normal? Screens:

1920mhz:

https://i.postimg.cc/RZPzz144/t1.jpg

And that 1800mhz for 1ms:

https://i.postimg.cc/cJmtWb66/t2.jpg


Of course its POWER LIMIT only in this test. So i guess its not faulty somewhere? In games i dont had any once such drops in clock.