Soldato
Ok, I’ll leave it for now. No intention of mining so I’ll just keep an eye (or ear) on the fans and see if they start ramping up more and more.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
30xx Series Founders Edition
- Thread starter Dirk Diggler
- Start date
Completely agree, fingers crossed.
Trust me the fans will let know when you hit the thermal limit, it's not a slight ramping up it's deafening. Even if you do it isn't going to damage your card. I read somewhere that the 110c limit was introduced in a driver update, so Nvidia obviously know about the "issue" and think it's fine. I only stress about the temps because I'm mining in my off time, otherwise I would just get on with gaming and forget about it.
Soldato
Its normally one of the two. Advice is to use the empty bog roll trick. Listen for the culprit.
GPu-Z shows the memory temps so quite easy to use that in conjunction with Afterburner to set a custom profile. Only problem is that the memory temps don't really cool down much even when you ramp fans to a noisy 80%+. Only way would be to add thermal pads or some extra fans blowing on the gpu. I get near 100C when playing some games like ARK or Cyberpunk with RT disabled
. I wish Nvidia had tried to keep memory temps below 90C but they skimped on the thermal pads which is forcing some to mod the cards themselves.
Soldato
Yep, now I have swapped them and can confirm a solid 15-20c improvement it is poor for such a scrimp when buying in bulk means the cost at factory to apply better pads would be pennies per card. The stuff you remove is horrible I wouldnt be surprised even with no mining just gaming that after two years it would be in good enough shape to keep the unit cool.
Soldato
Are you mining?
From what I've read, the pads they use are designed to maintain the same performance over the long term. Whether or not the pads people are using to replace them last as long, remains to be seen I guess. My past couple of times gaming for a few hours with HWInfo running has shown max VRAM temps of 90c. That's only playing EFT though which isn't hammering the GPU (around 40-50% utilisation). Something such as Cyberpunk 2077 may push the GPU harder.
Soldato
Yes.
I doubt in two years time thes stock pads will be performing as they did when they left the factory. Some spots you could peel and almost put aside to re-use, whereas others were a mush and disintegrated when you touched them.
Soldato
These things come with a 3 year warranty, surely they'll use parts they would expect to last for that duration.
Soldato
For some reason I thought it was two! Thanks
For the latter part, judging by the volume of people replacing them and seeing such an improvement either means:
a) the engineers/manufacturers know that the temps are tolerant and can withstand more abuse than people outside of that circle are letting on - or
b) the rush to bring them out and knowing the business profit margins meant although the pads are sufficient in one sense, could also be an oversight when/if it becomes an issue mid warranty lifecycle and they are deemed insufficient.
For the latter part, judging by the volume of people replacing them and seeing such an improvement either means:
a) the engineers/manufacturers know that the temps are tolerant and can withstand more abuse than people outside of that circle are letting on - or
b) the rush to bring them out and knowing the business profit margins meant although the pads are sufficient in one sense, could also be an oversight when/if it becomes an issue mid warranty lifecycle and they are deemed insufficient.
it seems that stressing over memory temps is a red herring. the reported memory temps are supposed to be the junction temps, not the chip surface temps which Samsung apparently specifies and which are lower. the driver starts throttling when the mem temps get too high - just like you can't kill the GPU die, you shouldn't be able to kill the mem either. of course if it starts throttling due to mem temps, then you have a problem.
think about it, would NV risk getting tons of returns inside the warranty period? of course their driver avoids that. they know people mine, and they can update their driver to cope with issues.
think about it, would NV risk getting tons of returns inside the warranty period? of course their driver avoids that. they know people mine, and they can update their driver to cope with issues.
Last edited:
Soldato
Isn’t the GDDR6X made by Micron?
I agree, they made an assessment that setting a 110c cap would save them enough money on claims to make it not worth putting better pads in, which would be very cheap and easy. Unless they're counting on hardly anyone having a warranty in the first place due to scalpers etc.
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 29,786
true.
Report: Why The GeForce RTX 3080's GDDR6X Memory Is Clocked At 19 Gbps | Tom's Hardware
"According to Igor, the general consent is around 120C before the GDDR6X chips suffer damage. This would mean that the Tjunction value should be set at 105C or up to 110C."
Soldato
So has anyone else found that installing the FE has pushed temps on system ram over it's thermal stability?
Permabanned
Have not noticed but now you mentioned it and I actually looked its blowing directly onto my RAM.
However I am only running at 3200 speed, its not overclocked just tighter timings.According to jay his system ran cooler with the card exhaust dumping air onto other components. Not sure how that works with hot air 0_o
Soldato
If the hot air is cooler than the components, it’ll still cool them, just not by as much.