Are these good stats for RX 6800 XT in Furmark burn in test?

[throwaway60540]

Found a used PC offer with the seller attaching an image of a burn in test in Furmark specifications below

 

Core: 2055MHz, mem: 1998Mhz, temp: 58c, GPU load: 97%, memory load: 95%, GPU power: 212W

FPS: 102 (min 101, max 106, avg 102), 2560x1080

 

Does this mean the GPU is in a good and stable condition?

[Tan3l6]

Depends how long the test has run. 58c on core is good, but again after how much time? And GPU hotspot temp should also be available?

Also - can't say if it's stable, just by fps and temps. Perhaps if it has run for an hour without glitches...

[RONOTHAN##]

16 minutes ago, throwaway60540 said:

GPU power: 212W

This makes it sound like they ran it with a lower power limit. The stock power limit for a 6800 XT is 285W, and Furmark is good at getting cards like this to run at the power limit consistently. That could be because they wanted to reduce the power draw of it, the cooler isn't performing adequately and it was the only way to keep it from overheating, or Furmark is misreporting the power figure for some other reason. 

 

Granted, Furmark is just about the worst test to determine if a card is good or not. It is about as far from a real world load as you can get, the performance metrics rarely make sense, and it's only good for thermal stress tests where you need a consistently high load. I'd ask to see a video of the card running something actually somewhat representative like 3DMark Time Spy to check for artifacting and compare the score to something else. 

[throwaway60540]

3 minutes ago, Tan3l6 said:

Depends how long the test has run. 58c on core is good, but again after how much time? And GPU hotspot temp should also be available?

Test ran for 20 minutes. Is that good enough?

[Tan3l6]

3 minutes ago, throwaway60540 said:

Rest ran for 20 minutes. Is that good enough?

Well, just by 20 minutes of suggested numbers it's not possible to say if it's stable, it's just that the seller is presenting numbers.

And @RONOTHAN## said the rest.

Still interesting would be to know how low are the "hotspot" temps if core is as low as 58C after 20 minutes, as that is really good.

[throwaway60540]

2 minutes ago, Tan3l6 said:

Well, just by 20 minutes of suggested numbers it's not possible to say if it's stable, it's just that the seller is presenting numbers.

And @RONOTHAN## said the rest.

Still interesting would be to know how low are the "hotspot" temps if core is as low as 58C after 20 minutes, as that is really good.

So, is it better to avoid this offer? Given that the card is running at a very low power?

[WereCat]

IDK how to judge performace based on Furmark bench... Furmark is usefull as a power virus to max out the power on the card as much as possible.

I can tell you that the stock power of 6800 XT is 255W and this card is definitely running in a power limited state.

 

In adrenaline SW you can go down to -6% from the stock power limit which would make it run at 239.7W max.

 

Considering this card runs even lower than that suggest that either it's further limited by max clock and heavy undervolt (which on stock is around 2500MHz) or has some other SW or HW modifications.

 

Either way, ask him to run this card in stock config and ask what are both the avg GPU temp and the hotspot temp.

[Tan3l6]

12 minutes ago, throwaway60540 said:

So, is it better to avoid this offer? Given that the card is running at a very low power?

Seller has most likely lowered "power limits", but why - no idea. 3DMark runs would be perhaps better proof, but also asking why the power limit is decreased might give some info imo.