Separate names with a comma.
Discussion in 'HardForum Tech News' started by FrgMstr, Nov 21, 2018.
Maybe it's NVIDIA's new subscription model, which includes an unlimited number of RMAs?
So that's how we ray trace our lives!
Well now that Micron vs. Samsung is out, one of the other plethora of rumors floating about this is anything using a NV reference board might be suspect.
Anyone heard anything about Strix 2080TI's?
There ain't no coup de ville hiding at the bottom of a Cracker Jack box.
It just works. . . .and then it doesn't!
Is it just the TI variants?
I don't know if that picture you posted on the article with the tombstone is your GPU, but it looks to me like it has protective plastic film
over the mid-section of the card, between the two fans.
Could that be?
If so, and the p[lastic film is still on the card, it wouldn't help in survival.
It's really bad no matter what that two random GPUs of the same family take a shit on day one of use.
I’m so glad the price of these cards turned me off from upgrading. What a fucking hassle it could have been.
I think your point enthusiast experience is really true. I’ve seen many youtubers bios mod, shunt mod, hacked on water blocks with chilled loops abuse their cards without failure. Hmmmm, wonder if Nvidia sent them cherry picked cards?
I was hoping the replacement card with a different bios was the answer to these failures. Clearly by now Micron is not the culprit. I guess as I suspected the board design is at fault aside from multiple other quality control fiascos.
Thank you for your continued coverage on this.
No heat is radiated through that portion of the card of any consequence. If you are suggesting that is the reason for failure, you are WAY off base.
Wow, the heatsink in this pic looks like its about to melt of the damn thing. Cool Pic.
Read verified reviews about the 2080 and 2070. They all do space invaders when given the chance. Luckily my MSI 2070 Gaming Z only had one TDR so far. I have been benchmarking the crap out of it to make sure its running OK while I am in return period.
I think, if nothing else, we've proven their .01% RMA number is absolute bullshit. Compared to the number of 2080ti cards purchased by [H] members, and the number of failures...
Just with Kyle's cards and whoever said both of theirs died...that's rolling a .0001 4 times out of 5. Stats tells me that's something something standard deviation something sin cosin tangent pi squared - fucking unlikely.
All I want for Christmas is no one to get me a 2080ti. I don’t need the headache. If I was looking for a risky short term investment I’d buy whatever-is-cool-this-week crypto coin.
Sure it might be worthless in a week or two, but at least it doesn’t catch on fire or invite space invaders to my house.
I think I am going to point the issue at IMC degradation...
I smell a recall or class action coming if Nvidia can't shake this wave of defects.
Obviously that number is BS. From https://hardforum.com/threads/do-you-have-problems-with-your-rtx-card.1970788/page-7#post-1043947404 there are about 20 people who've stated their card has died. That would mean that this site would need roughly readers 200000 people who have a working 2080 TI and didn't bother to vote, which I don't think is realistic.
I doubt its a design flaw in the GPU/IMC itself. I think the problem is heat. Look at the pick above that Kyle took. The middle area of the heatsink is approaching 90 C. There is no way any of the memory chips can handle this heat at longer stretches of gaming / benchmarking. It also explains why AIB cards even with a NVIDIA PCB last longer and the custom designed cards are fine.
Did you try installing the latest Radeon drivers?
There are custom cards out there with custom PCBs doing space invaders. It is not just the FE cards. The FE cards seem to have more population at this time and I am certain that others are starting to become a little gun shy even purchasing one at this point. I sat at BestBuy and stared at the box for 15 minutes before I bought it.
Same PSU for all your failures?
That's all I got.
Rgr, the custom PCB cards are running in the mid 70's at default memory speed. Most people run them overclocked with the oc scanner which means high 70's to low 80's. Its still about 15 to 20 C too high. Its only a matter of time.
AMD is like "can't beat us in benchmarks if your cards don't work"
This is starting to remind me of the ol Xbox 360 RROD
hmm just thought of something.. nVIDIA MAY BE GOING THE WAY OF M S + WIN 10 .....GPU'S AS A SERVICE ?????
ie you need to keep replacing them ..smart move Ngreedia ??????????????????/ at $1200 bucks a card you get into more influx o cash for the end of the yr sales .. even if you screw up the drivers ..and cook cards ..
and RTX and so called ray ray traced games fail ..smooth move..lets jack prices ..
I just can't help but to pause for a moment to wonder after Lisa Su read this, and then took a peek at nVidia's stock prices today (they are all way down to $144 from $293 few days ago) to see if there's any impact there, just how that went...
Calm down and take a deep breath.......
These pictures are taken on an open air test bench with the ambient about ~72F. This is with Heaven running in a loop after heart soaking the card. I was saving these for the article, but that is not happening for a week or so now most likely. The first picture is with the backplate removed.
Great pics, if these temps are stock clock than wow at the one without the back plate. This is some creazy FLIR prOn. Thanks again !
Stock clocks running Heaven at 1440p.
Step 1...place RTX on sheet of paper.
Step 2...hire a guy named Raymond.
Step 3...task him with having to run a crayon around the perimeter of the RTX, transferring its outer shape onto the paper.
Step 4...tape the paper onto your computer monitor.
And there you have it: Ray Tracing that'll actually work.
Wow that sucks. I think the bin for bad 2080ti gpu is far higher than nvidia leads on.
That is pretty bad luck, if I would even call it luck per se. It reeks of bad quality control at Nvidia and company....
Seems we havent seen the last of this yet, I wonder how many more will go. I wonder if Nvidia has ACTUALLY figured out what is going on, hopefully not a fault in the silicon, resale on the rev 1 2080Ti FE is gonna down the shitter thats for sure
Wellllllllllllllll.... that explains the whole NDA thing now, doesn't it?
Heaven looping. Updated post.
I nearly pissed my pants laughing at the last line re: the nda.... hilarious.
doubt it's the gpu it's self, personally i wouldn't be surprised if the cards/circuitry/vrm's/etc were designed for the gpu to be running at a lower clock speed/voltage then what they ended up releasing it at due to performance being too close to the 1080ti. but we may never know exactly what the cause is.
nope and nope.
I can't believe the 5 year blanket NDA