GeForce RTX 3080 CTD Issues likely due to POSCAP and MLCC configuration

This will end in a lawsuit.

3080 owners will get $1.09

 

I'll just quote TiN on this, "I am not convinced at all that culprit is in using few MLCC or polymer cap behind a GPU, it's just misinformation and guesswork spread by one blogger IMHO."

Also calling them Poscaps is like calling every CPU Ryzen or.. i5, or every car a Fiat

 

This exact card is not mentioned by several sources for no reason. Your setup, silicone, environment and usage is different from others.

This is what I am thinking, and most afraid of. I mean if the TUFF version have 6 MLCC's and there are reports of this card also CTD´ing, then the capacitors might not be the full story.
I did hear a rumor that reviewers got a different setup than normal TUFF cards, but cannot verify.



It's really getting complicated as a customer to choose right now, since you not only want the one having great performance, but also stability. This really adds more headaches, as nobody is 100% sure of anything, which makes any purchase a risk. Hopefully this will all be more clear when more and more people get their hands on all the different brands, and extensive testing have had time to complete.

 

Another scandal:

https://www.reddit.com/r/nvidia/comments/ixvcmu/received_3080_from_neweggca_just_the_gpu_no/
https://www.reddit.com/r/nvidia/comments/iyir8j/this_is_how_my_card_came_no_box_newegg_plate/

Bent plate:
https://preview.redd.it/7z38ralcmyo...bp&s=5373f7ce9665c2a741b27c63165532b5b3fc0c71

And so on...

 

That's no scandal and not an nvidia gpu specific thing, blame your retailer.

 

Yeah, POScaps (according to BZ) is just a common industry used term for this general variety of caps, but are rare and not actually used for this purpose.

 

And you have the breakdown of who is affected and who is not?

 

Everyone with some knowledge can call this thing a BS. Reason for CTD's and other instabilities are elsewhere. NOT THERE people are currently looking. Misleaded and wrong info made a confusion all over the net. The power of internet is a bomb for noobs.

 

Nvidia's problem is that it didn't allow the board partners to test their cards properly, only with the extremely limited testing software Nvidia provided, and which didn't, in fact, test the cards properly. In ideal world the partners should be able to get compensation from Nvidia, but like I said in another post, Nvidia is a bully to work with due to its overpowering market dominance (and Mr Leather Jacket's personality), so if a board partner complains, I reckon Nvidia will instead cease working with that particular company or perhaps delay any future chip shipments to it, giving its competitors an edge. Why did this all happen? So that Mr Leather Jacket would have a much nicer announcement video, where no leaks had spoiled his numbers.

The issue is not whether it clocks up or not beyond the advertised value, the issue is crashing to the desktop.

More like VW promising their cars fulfill the emission standards, but in reality they don't. So, updates would need to alter the engine performance in order to do it, meaning it's not anymore the car the customer bought. Just like Apple altered phone performance so that an old battery would last longer. The customers can return the cards normally using the warrantly and get a new card (that might also not work) or get their money back. Perhaps if you return your card for warranty, they will just reflash a lousier bios and return the same card to you. It would still work within the advertised values, but it would be a lesser card. That's why I'd think it wiser to try to get your money back instead. Waiting some months might see new revisions of the card with actually working hardware and thus better boost clocks without harmful bios alterations.

 

https://twitter.com/HardwareUnboxed/status/1309659834468298753

 

I hope that tomorrow Nvidia relases an statement of this issue and add some information about it, because today all is rumour and noise... if FE are having same problems, I want to doubt that this is a HW problem...

 

Haven't heard about FE's having these problems other then from random people on the internet with no proof and are likely just trolls.

 

It's possible, but also the CTD issues come from some people on the net, and none of the reviews websites had this problems...

 

i haven't read or heard of any FE cards having the CTD's but that might just be because I haven't heard of anyone who actually has one.

 

Gigabyte is using 6 sp-caps with 470 microfarads under the core of the 3090 Gaming OC (also in their Eagle OC), while other companies like Zotac is using 6 units of 330. Colorful is also mounting 330 in the 3080 as seen in Jayz's video.

https://www.techpowerup.com/review/zotac-geforce-rtx-3090-trinity/images/back_full.jpg
https://www.techpowerup.com/review/gigabyte-geforce-rtx-3090-eagle-oc/images/back_full.jpg

My card is a Gigabyte 3090 Gaming OC, I have shunt modded it, not a single crash playing BFV for hours, I also let the card running with Heaven 3 hours at 2025-2055 1.062v-1.1v without a problem.

 

well evga said it wasn't stable.

 

Hardware Unboxed:

''The crashing with the RTX 3080 cards doesn’t appear to be down to the caps used, which is why we haven’t made a video yet, we don’t know the issue. What we do know is the FE and TUF Gaming models crash just as much as other models and they use MLCC’s.''

Just added to my confusion..

 

no they don't.

 

I don't know, then why would they claim that FE and TUF Gaming models crash too, I can't test more than 2 GPUs, unlike them..

 

People are fanboys and will claim whatever they can to make the "other side" look bad. Maybe some TUF cards are crashing, i dunno, i've seen some articles talk about that, but haven't heard anything about the FE.