AMD RX Radeon 7900 XTX (reference) Failure rate would be 11% opposed to 1%

Published by

Click here to post a comment for AMD RX Radeon 7900 XTX (reference) Failure rate would be 11% opposed to 1% on our message forum
https://forums.guru3d.com/data/avatars/m/248/248994.jpg
This is a worse failure than Nvidia's melting power adapters, even if not as dramatic sounding. Scott Herkelman probably has been called a few times by Lisa Su for explanations.
https://forums.guru3d.com/data/avatars/m/216/216349.jpg
More than 10% is too much.
https://forums.guru3d.com/data/avatars/m/282/282473.jpg
"jebaited"
https://forums.guru3d.com/data/avatars/m/181/181421.jpg
I suspect many of these RMA are not faulty cards but customers that don't want to take the risk to keep a perharps faulty one. I would do the same.
https://forums.guru3d.com/data/avatars/m/258/258664.jpg
Wait a second...
RMA instances are 9–11% per thousand of cards, although they should be less than 1%; several AMD partners (AIB) are reporting the failures.
Does that mean the partners that manufacture the MBA cards, or do 3rd party coolers now act up as well? o_O edit: source (and google translate) help: It's about OEM partners, actually, who should have built rigs with MBA cards. Also, source says that Cooler Master made the cooler, but the vapor chamber was done by a company named "Asia Vital Components" (AVC). Apparently, AVC's quality control failed.
data/avatar/default/avatar38.webp
Kaarme:

This is a worse failure than Nvidia's melting power adapters, even if not as dramatic sounding. Scott Herkelman probably has been called a few times by Lisa Su for explanations.
How is it worse? One failure ended up with a melted mess and possible failed card, the other throttles a bit. I presume both manufacturers will be honouring RMAs
https://forums.guru3d.com/data/avatars/m/248/248994.jpg
bnauk:

How is it worse? One failure ended up with a melted mess and possible failed card, the other throttles a bit. I presume both manufacturers will be honouring RMAs
Didn't Nvidia suffer something like 0.05% of product returns, whereas AMD is suffering a whole lot more? How is that not worse? The Nvidia case only looked worse on the surface because there was smoke involved. Not to mention in Nvidia's case the primary reason was user errors encouraged by the less than thoughtful design, whereas in AMD's case it's verified faulty products. Nvidia could simply stop most of new cases by spreading information, AMD can't do anything but accept the products back for a replacement or refund.
fantaskarsef:

Also, source says that Cooler Master made the cooler, but the vapor chamber was done by a company named "Asia Vital Components" (AVC). Apparently, AVC's quality control failed.
Alas, quality control is the most risky thing to outsource to the contract manufacturer won through the lowest bid. Not to mention the one whose brand adorns the product is responsible for the quality, not whoever physically manufactured it wherever. Both AMD and Cooler Master should know better.
https://forums.guru3d.com/data/avatars/m/108/108389.jpg
AMD didn't win the performance crown, but at least they won the RMA award this generation πŸ˜€
https://forums.guru3d.com/data/avatars/m/248/248291.jpg
Krizby:

AMD didn't win the performance crown, but at least they won the RMA crown this generation πŸ˜€
You can't write karma, without RMA.
https://forums.guru3d.com/data/avatars/m/258/258664.jpg
Horus-Anhur:

You can't write karma, without RMA.
fml that's deep, nerd humor post of the day. πŸ˜€
https://forums.guru3d.com/data/avatars/m/295/295892.jpg
Who buy AMD cards after NV 40 series anyway? Who need headache? NV FTW and coming the new AI Driver with 30% performance boost!! OMG!! 😎
https://forums.guru3d.com/data/avatars/m/246/246171.jpg
Wault:

I suspect many of these RMA are not faulty cards but customers that don't want to take the risk to keep a perharps faulty one. I would do the same.
If I bought one, I'd rather keep it and run some stress tests just to double-check, especially since AMD seems to care about the exact results. I'm too lazy to go through the RMA process if I don't have to. If I care enough to spend close to (or over) $1000 on a GPU, I obviously care enough about gaming that I'm not going to deprive myself of it for a what-if scenario, as anyone who is experiencing thermal issues is bound to know it. But, that's just me. I also would never spend that much on a GPU. EDIT: In a cynical perspective, perhaps AMD does actually have enough stock, but they're telling people otherwise because they're hoping it'll either discourage people from RMA'ing it (since it means they'd be on a waiting list) or it'll entice people to attempt to fix the problem themselves, thereby voiding the warranty.
https://forums.guru3d.com/data/avatars/m/248/248994.jpg
Stratos2k22:

Who buy AMD cards after NV 40 series anyway? Who need headache? NV FTW and coming the new AI Driver with 30% performance boost!! OMG!! 😎
Technically 4070 Ti should have been the upper limit of my interest, but Jensen decided to smoke some really, really potent stuff (probably borrowed from Elon '$44B Twitter' Musk), and it costs over 1000 euros. Thanks, but no thanks. Depending on the prices of the actual mainstream with more sensible prices, AMD could still be getting my money. If the mainstream cards ever deign to be released by either manufacturer.
https://forums.guru3d.com/data/avatars/m/152/152580.jpg
Kaarme:

Didn't Nvidia suffer something like 0.05% of product returns, whereas AMD is suffering a whole lot more? How is that not worse? The Nvidia case only looked worse on the surface because there was smoke involved. Not to mention in Nvidia's case the primary reason was user errors encouraged by the less than thoughtful design, whereas in AMD's case it's verified faulty products. Nvidia could simply stop most of new cases by spreading information, AMD can't do anything but accept the products back for a replacement or refund.
AMD's problem is a manufacturing problem and in the worst case, it can lead to a performance degradation of the card. NVidia's problem is a design problem, and in the worst case, it could lead to the building burning down and the loss of life. NVidia's explanation of "user error" is nonsense. Never have graphic cards (or any other PC component) required special training to be properly connected to prevent the possibility of starting a fire. NVidia's problem has not been solved in any way, because computer vibrations can sooner or later lead to loosening of the link and melting of power sockets and pcb. Personally, I would not dare to leave the computer with the RTX 4090 with a demanding game turned on for the night (e.g. in order to farm a skill) and go to sleep. Regarding statistics, only NVidia knows how many cases have been reported and the numbers they quote may be or not be true. The problem is that after the NVidia scandal broke, there was an immediate wave of defenses from various "Igor experts", which prevented a thorough investigation of the issue, unlike the AMD scandal, where the "Igor experts" quickly found their anonymous sources of information, which immediately evolved on the Internet from a few to several hundred and then thousands of cases.
https://forums.guru3d.com/data/avatars/m/248/248994.jpg
leszy:

AMD's problem is a manufacturing problem and in the worst case, it can lead to a performance degradation of the card. NVidia's problem is a design problem, and in the worst case, it could lead to the building burning down and the loss of life. NVidia's explanation of "user error" is nonsense. Never have graphic cards (or any other PC component) required special training to be properly connected to prevent the possibility of starting a fire. NVidia's problem has not been solved in any way, because computer vibrations can sooner or later lead to loosening of the link and melting of power sockets and pcb. Personally, I would not dare to leave the computer with the RTX 4090 with a demanding game turned on for the night (e.g. in order to farm a skill) and go to sleep. Regarding statistics, only NVidia knows how many cases have been reported and the numbers they quote may be or not be true. The problem is that after the NVidia scandal broke, there was an immediate wave of defenses from various "Igor experts", which prevented a thorough investigation of the issue, unlike the AMD scandal, where the "Igor experts" quickly found their anonymous sources of information, which immediately evolved on the Internet from a few to several hundred and then thousands of cases.
You could be right, but at the end of the day you are still only dealing with could bes. As far as we know, not a single house has caught fire because of the connector. All we have seen are connectors damaged by heat, that's it. The connector also has a locking mechanism. The problem is that some users didn't dare to use enough force to push the plug all the way in, and thus the locking latch wasn't engaged. It's a weakness in the design, yeah, I've been saying that myself from the beginning, even if nobody in the industry wants to assume responsibility, but technically it still works if used properly. However, there's nothing you can do about the flaw in the AMD cooler if you happen to receive a card with the issue.
data/avatar/default/avatar26.webp
Wault:

I suspect many of these RMA are not faulty cards but customers that don't want to take the risk to keep a perharps faulty one. I would do the same.
They would deny that false RMA, it is going to show up in the test run and the owner is going to get billed. I could not even RMA my ASUS Vega 64 with 110 degrees VRM temp, they send me a AMD statement that said 110 C hot spot was ok. I replied that my hot spot was 80 C and ok, but the problem was the Asus designed VRM cooled by the Asus designed cooler. The reply was that it was working as intended.
data/avatar/default/avatar34.webp
fantaskarsef:

Wait a second... Does that mean the partners that manufacture the MBA cards, or do 3rd party coolers now act up as well? o_O edit: source (and google translate) help: It's about OEM partners, actually, who should have built rigs with MBA cards. Also, source says that Cooler Master made the cooler, but the vapor chamber was done by a company named "Asia Vital Components" (AVC). Apparently, AVC's quality control failed.
This problem only affects the reference design, also known as Made By AMD (MBA) cards. AMD controls the manufacturing of these cards, but they are packaged and sold under various AIB brand names (such as Sapphire, Powercolor, XFX, etc., this is not about PC builders.) Because of this, there are faulty cards being sold under every brand that offers MBA cards (I'm not sure if all of them do) hence the statement made.
https://forums.guru3d.com/data/avatars/m/266/266726.jpg
its shouldnt cost amd that much other than bad rep, thankfully for them the cards can simply be fitted with a new cooler and resold. it will cut into their margins a bit ,but since the cards are already priced high the financial damage will be minimal imo.
data/avatar/default/avatar33.webp
Kaarme:

Technically 4070 Ti should have been the upper limit of my interest, but Jensen decided to smoke some really, really potent stuff (probably borrowed from Elon '$44B Twitter' Musk), and it costs over 1000 euros. Thanks, but no thanks. Depending on the prices of the actual mainstream with more sensible prices, AMD could still be getting my money. If the mainstream cards ever deign to be released by either manufacturer.
I agree with you with Jensen but please do some homework, Elon Twitter deal was done to expose the crooked F B I and our disgusting administration. It wasn't about the money. After it was exposed many Twitter employee's came forward and so did a few F B I agents. James Woods came forward and showed evidence of being harassed by the previous Twitter ownership and false allegations which ruined his acting career and now is suing and has a good chance of winning. You should be thanking Elon since it proved how Twitter band a large amount of doctors that discussed the truth about C O V I D. Prove me wrong and I will give you $1,000.