Is this you Peachy?

1337welshe
Is there some sort of company that does certified graphics card repairs?
 
My gpu works, and games fine.. And by fine i mean underclocking everything to the absolute minimum and running the lowest possible game settings, and unless ive done that the driver will crash or the game will crash or the entire pc will lockup, i believe the VRM is becoming to hot and then crashing if it has high load.
And why i believe its the VRM is because i repasted it twice, reconstructed it, and tempatures are completely within safe margins, but to touch this thing is like a bombfire on steroids..
 
Any possible help would be great (i don't want to repair it myself, if you even can..)
 
Edit : forgot to add that my gpu is a EVGA 980Ti - ACX 2.0 Superclocked

While it is usually the GPU, have you considered your PSU might be at fault? 70-80C is very hot to the touch.  The GPU is designed by default to run around 180F or 80-82C under a full load. The VRMs are typically rated to 105C. Too answer your question directly no, there are no official certified repair facilities for Nvidia GPU's. When you under clock,  can you get stability just under clocking  core or memory or does it require both. If both, I am more suspicious of your PSU.

HeavyHemi

1337welshe
Is there some sort of company that does certified graphics card repairs?
 
My gpu works, and games fine.. And by fine i mean underclocking everything to the absolute minimum and running the lowest possible game settings, and unless ive done that the driver will crash or the game will crash or the entire pc will lockup, i believe the VRM is becoming to hot and then crashing if it has high load.
And why i believe its the VRM is because i repasted it twice, reconstructed it, and tempatures are completely within safe margins, but to touch this thing is like a bombfire on steroids..
 
Any possible help would be great (i don't want to repair it myself, if you even can..)
 
Edit : forgot to add that my gpu is a EVGA 980Ti - ACX 2.0 Superclocked

While it is usually the GPU, have you considered your PSU might be at fault? 70-80C is very hot to the touch.  The GPU is designed by default to run around 180F or 80-82C under a full load. The VRMs are typically rated to 105C. Too answer your question directly no, there are no official certified repair facilities for Nvidia GPU's. When you under clock,  can you get stability just under clocking  core or memory or does it require both. If both, I am more suspicious of your PSU.

Yeah i absolutely have. Fortunately my motherboard + memory and cpu were at the time this started to happen under warranty. Immediately i rma'd all my components and to make sure it wasn't my PSU i waited for my motherboard + cpu + memory to come back after being fully tested, and then purchased a brand new power supply unit, unfortunately this did not solve anything but made me find more problems, IE.. "HDMI" no longer works at all on my 980ti and shows red lines all over the screen but display port (which i use now..) has 0 issues apart from the game crashing, locking up pc etc, I also used multiple different windows drives with no backup drives to ensure it wasnt my hard-drive / ssd at fault as well, i came to the conclusion after every single component was replaced other than my GPU, but because i don't have a spare GPU i can't really test if another gpu fixes the issue, and i don't have the means to afford to buy something else i would much rather this one be repaired if possible ^^
 
And to answer about the underclocking, kind of.. This 100% helps with stability, but high gpu load games will immediately crash to desktop (usually throwing up a game error) or it will freeze my entire PC, I even did multiple checks on the voltages (nothing in my PC has ever been overclocked or touched in regards to getting more performance (obviously my gpu is overclocked) but its factory and stayed that way.
 
As for Precision X settings that add maybe 30% more stability to my system these what are what i need to change :
-201 memory clock, -90 clock and then it becomes slightly more stable and actually allows me to play some games, BUT games will crash its just the question of when.. It becomes random..
Games like "Ark Survival Evolved" crash within 4minutes every single time regardless of game settings.
Games like "PlanetSide 2" crash within 1hour and 52minutes (it varies on this game. Sometimes 5mins sometimes close to 2 hours.)
 
Also increasing the fan speed to 100% it stays WELL within safe tempatures barely touching 60C most of the time (underload) but this makes the stability even worse.
 
Adding more fans aiming my GPU also does not make it more stable, infact it makes it worse and crashs more often.
 
Thanks, Sorry for the essay!
 
"Also".. Forgot to add, i DID rma my gpu when it first started happening, via the vendor that sold me the GPU to be told the GPU is completely fine and there are 0 issues. But i beg to differ considering ive replaced every single component, without my GPU stress testing and games run completely fine on my on-board graphics. So its 100% my gpu

I wonder if Evga is providing also out of warranty paid support,
Both of my 980Ti Classifieds died - one like 3 months before warranty and this one was replaced but second I think 2 months after.
I'm now regretting not taking extended warranty but I never thought those cards will still provide enough performance in 1440p after 3 years.
One card is obviously not enough for newest titles and we all know the availability of new 30x0 cards :(

Lothar_ab
I wonder if Evga is providing also out of warranty paid support.

No

If someone built a company to repair old video cards I think it would be financially sound. 

ty_ger07

Lothar_ab
I wonder if Evga is providing also out of warranty paid support.

No

Short and to the point. 

Lothar_ab

Puddles.

Lothar_ab
I wonder if Evga is providing also out of warranty paid support,
Both of my 980Ti Classifieds died - one like 3 months before warranty and this one was replaced but second I think 2 months after.
I'm now regretting not taking extended warranty but I never thought those cards will still provide enough performance in 1440p after 3 years.
One card is obviously not enough for newest titles and we all know the availability of new 30x0 cards :(

The 980ti's were amazing cards for their time, specially the EGVA Hybrids. Mine too was still going strong right up until 2 months ago. Seems to be a LOT of 980ti's failing around now for some reason (tin foil hats).  As for paid out of warranty support, I'm going to have to go with a cautious no on that one. I've been trying to get a response from EU Customers Services for the past few weeks. Yesterday I received the standard: 'Sorry, your card is out of warranty line.'  Which is fine but proves to me, like many companies customers services now, it's become a binary script response service who don't read emails properly. My email included the point that I'm aware my card is out of warranty and in fact I was looking  to purchase one of their listed B stock 980Tis (The EU store has no purchase buttons at all) or any old stock if they did not offer a paid repair service. Seems you can't even give money to EVGA now as the response was: your card is out of warranty. < took them two weeks to give me that response too. So I wouldn't get your hopes up on rescuing those cards. Be nice if we could though. It would mean I could get my PC back up and running. 

I'm a really big fan of Evga mainly because of the uniqueness of their products and outstanding support but unfortunately, this brand is almost dead in Europe.
I even read on this forum that somebody who was able to recently order a card through Evga EU online shop, the product has been shipped from Singapore and he had to pay additional tax. This is ridiculous.
 
It's a pity because it's very hard to find a trustworthy brand nowadays. I still have some small hope for their resurrection though but it might take a while due to the current situation in the world.

 
Sadly this is true of many Brands in the EU at the moment, specially in the UK. EVGA is one of the brands that I held up there for bending over backwards to help their customers. But my recent interactions with them just come across as heartless! 

HeavyHemi

rjohnson11
If someone built a company to repair old video cards I think it would be financially sound. 
 
The retro market MIGHT be willing to pay, but for the average customers its' not financially viable. It's a throwaway society.

Considering the warranty periods and the pace of technology, I just don't see this happening...Also you would need access to proprietary non public info 'schematics' and there must be a reason it has not in some 4 decades.

This is part of the reason it's not financially viable. Without schematics it becomes a lengthy and costly affair do diagnose. A hobbyist may have a enjoyable time trying to repair something, but it would be too time consuming to do as a business without the relevant information.

Lothar_ab
Also many times he replaces the broken element and after few months another one breaks - it's like every fault have some kind of impact on the rest of the small elements.

True. Every component is rated for a finite lifetime, and for profitability, the components are chosen to be the least expensive component version available which will at least exceed the intended warranty of the product.  Having components fail after the warranty expires is to be expected.  Having multiple components fail after the warranty expires should be similarly expected.

Puddles.

ty_ger07

Lothar_ab
Also many times he replaces the broken element and after few months another one breaks - it's like every fault have some kind of impact on the rest of the small elements.

True. Every component is rated for a finite lifetime, and for profitability, the components are chosen to be the least expensive component version available which will at least exceed the intended warranty of the product.  Having components fail after the warranty expires is to be expected.  Having multiple components fail after the warranty expires should be similarly expected.

Designed Obsolescence.
 
Just letting my brain fart here. This would infer that said components would be chosen for the longest warranty period offered on said product, in the OP's case, a EVGA 980Ti. Now I may be mistaken here so please correct me if I am wrong, but I believe the paid optional extended warranty available on these cards was for 5 and 10 years? It then stands to reason that the components on these cards were optioned to last at least ten years before failures should be expected? Yes. Not 5? Following that thinking through would suggest components should be available for repair for at least 10 years past last production date. Many other industries support legacy products, automotive for one, but the electronics industry is one of the largest offenders for not doing this. Huge cases are being fought over the right to repair at the moment, with the whipping boy being Apple for an industry wide problem. Is it any wonder that there is a HUGE problem with E-waste right now if we nurture a culture of 'it's dunski, throw it'.
 
 
Steps of box....
 
 
 

Not really. The only 'unique' design component of the GPU is essentially the die. Everything else is pretty much a commodity item with long established MTBF. The issue here is recycling something of this complexity profitably. Back in the day when I was running the ATE division for one of the largest semi firms on the planet we recycled millions in raw material from defective products. However, at this point in the process they were not attached to the 95% waste product... the PCB and other packaging. No laws regarding repair are going to reduce the complexity and the difficulty in repairing and sourcing materials or items of this complexity.

ty_ger07

Puddles.
This would infer that said components would be chosen for the longest warranty period offered on said product, in the OP's case, a EVGA 980Ti. Now I may be mistaken here so please correct me if I am wrong, but I believe the paid optional extended warranty available on these cards was for 5 and 10 years? It then stands to reason that the components on these cards were optioned to last at least ten years before failures should be expected? Yes. 

No.  The extended warranty costs extra because the amount of failures which will occur within the 10-year warranty period increase compared to the standard warranty.  If all the components EVGA used were expected to last 10 years, EVGA wouldn't make the 10 year warranty an "extended" warranty which costs extra.  Honestly, I would expect nearly 100% of the cards will fail within the 10 year extended warranty, if they were used constantly.
What EVGA bets on is that most people will use the card only periodically and not primarily at full load.  It also bets that most people will run out of use for the outdated card long before the 10 year extended warranty expires, and thus never use the warranty.

Where the 10 year and former lifetime warranty help out is if you can get a useful newer and more efficient replacement. You'd be surprised how low the 10 year failure rate is. Less than 10%. Around 90% are retired while still functioning. You might be surprised how ancient some of the test equipment the electronic industry still uses for measurement and testing. Of course we cannot say what percentage of those now collecting dust would still work if fired up. One the the largest failure mechanisms is reapplying power after a significant idle time as all caps and components have bled down to a zero state. DC caps degrade while idle and can cause an inrush current until they reform their dielectric oxide layer after application of voltage.

HeavyHemi
 
 
Not really. The only 'unique' design component of the GPU is essentially the die. Everything else is pretty much a commodity item with long established MTBF. The issue here is recycling something of this complexity profitably. Back in the day when I was running the ATE division for one of the largest semi firms on the planet we recycled millions in raw material from defective products. However, at this point in the process they were not attached to the 95% waste product... the PCB and other packaging. No laws regarding repair are going to reduce the complexity and the difficulty in repairing and sourcing materials or items of this complexity.

 
Valid point on the GPU die. Putting that aside, as you say, most decent components now days have a long MTBF and one would hope that EVGA, with their long standing reputation for quality, wouldn't skimp on bean components in order to build in obsolescence. One would assume that the 5 year or even 10 year transferable warranty, paid or otherwise, would be based on the confidence that their products have been designed to go that distance under normal use with, hopefully, low percentage of failures.  Now as for recycling, I'm one hundred percent with you. The complexity of these items has got to the point where recycling has become somewhat challenging, and IMHO this is where things need to change. Companies could design products to be more easily recyclable or repairable, but they don't. Repair shops have a difficult time obtaining schematics; propriety systems are becoming more common place rendering products unrepairable, and items are assembled in such a way that they are impossible recycle, let alone repair. This is what the court cases are about, give the people the tools to have a choice to repair, not lock everything behind walls. Recycling is a whole different court case.......Across the hall.
 
Get's down off box again 

No worries my ol' china. Thank you.