Overclock.net banner

1 - 20 of 59 Posts

·
Expert pin bender
Joined
·
2,185 Posts
Discussion Starter #1 (Edited)
https://www.extremetech.com/gaming/313458-reseller-rma-data-shows-fascinating-pattern-between-amd-nvidia-gpus

Here are the high-level takeaways from the chart, in no particular order:

Less-complex, less-powerful GPUs fail less often than more complex, more powerful GPUs.
AMD’s midrange and budget cards do not fail more often than midrange and budget cards from Nvidia.
PowerColor AMD GPUs fail more often than other brands.
The RTX 2080 Ti is the GPU statistically most likely to fail. It is the only GPU with two-digit failure rates (11 percent) reported from multiple vendors.
AMD high-end GPUs fail more often, in absolute terms, than Nvidia GPUs, even if we remove the impact of PowerColor from the AMD data. The gap is significantly smaller if you do, however.

any surprises? What is amd's high end anyways... like Nvidia is high end and AMD is low end at best.

EDIT: adding a better link thats shows how bad AMD is

https://www.techspot.com/news/86236-retailer-data-shows-amd-5000-series-has-almost.html
 

·
PC Evangelist
Joined
·
46,731 Posts
Lol. One is 2% other is 3%. 50% higher.
 

·
Premium Member
Joined
·
6,318 Posts
Lol. One is 2% other is 3%. 50% higher.
I was definitely happy to be reading some info we don't get to see everyday, However, this. its like zoomed in graphs in text as marketing somehow.

Info good, exaggerations not welcome.
 

·
professional curmudgeon
Joined
·
10,391 Posts
idk, i'd be giving power color the stink eye if i was amd:

HERE:

Capture.PNG
 

·
WaterCooler
Joined
·
3,486 Posts
Interesting to see the brand/model RMA breakdowns.
 

·
mfw
Joined
·
8,643 Posts
Note to self: don't ever buy PowerColor.
 

·
I <3 narcissists
Joined
·
6,223 Posts
It's interesting data, would be more interesting if we could get it from more retailers to get a larger pool. It explains why I have never had to RMA a graphics card, percentages are pretty low.
 

·
WaterCooler
Joined
·
3,486 Posts
One does wonder if the higher AMD RMA rates for the 5000 series were more related to the broken driver state those cards sat in for a while versus the actual cards going bad?

EDIT: Also looks like we're missing some brands in this list. Noticed EVGA wasn't on the Nvidia lists, and don't see ASUS either.
 

·
Performance is the bible
Joined
·
7,050 Posts
Not sure I understand how people return x.y cards. Is it someone returning only the cooler saying "yeah, the rest of the card went up in flames"?
Someone rounded the wrong number it seems.

It would also be interesting to see if cards like the 2080 TI were also more damaged because people were trying to push them too much over actual hardware failure.
I would expect people to try and push a 2080 TI more than they would a 5500 XT. Also wonder the margins of errors on the OCed versions of cards.

But overall, even with the 2080 TI going to 11% failure rates on some brands, it is amazing to see how many people bought the 2080 TI. And people claimed that those cards aren't selling as much, when being so high priced.

Also how little in comparison the low end is selling. My guess 2nd hand cards are taking a bigger cut out of the low end range than we expect.
And how big the mid range is.

One does wonder if the higher AMD RMA rates for the 5000 series were more related to the broken driver state those cards sat in for a while versus the actual cards going bad?

EDIT: Also looks like we're missing some brands in this list. Noticed EVGA wasn't on the Nvidia lists, and don't see ASUS either.
My guess is because both asus and evga have their own program in the US. You can get a replacement card directly from them instead of going through the vendor you bought it from.
Or maybe the vendor doesn't sell those brands.
 

·
ٴٴٴ╲⎝⧹˙͜>˙⧸⎠╱
Joined
·
6,083 Posts
2080ti owners more critical to get samsung memory ayy?


return micron since it has an overclocking defect :^)
 

·
Registered
Joined
·
616 Posts
idk, i'd be giving power color the stink eye if i was amd:

HERE:

View attachment 364170
Tbf given the sample sizes involved it could literally just be a few bad batches or somebody in a warehouse practicing juggling with a stack of 2080Ti's before they ship out.

Certainly interesting but not a large enough dataset to look statistically significant.
 

·
Registered
Joined
·
3,862 Posts
It's interesting data, would be more interesting if we could get it from more retailers to get a larger pool. It explains why I have never had to RMA a graphics card, percentages are pretty low.
I had to RMA 3 AMD 590X gpu's in a row. I believe they were powercolor and sapphire.
 

·
Why is it smoking?
Joined
·
4,006 Posts
It's interesting data, would be more interesting if we could get it from more retailers to get a larger pool. It explains why I have never had to RMA a graphics card, percentages are pretty low.

The only GPU I have ever had to RMA was a Gigabyte 5870 which would artifact at stock clocks. It's also the first (and only) ATI/AMD card I ever purchased that wasn't either AMD branded or a Sapphire. I know a sample size of 1 is not data, but Sapphire has always been good to me.
 

·
professional curmudgeon
Joined
·
10,391 Posts
Tbf given the sample sizes involved it could literally just be a few bad batches or somebody in a warehouse practicing juggling with a stack of 2080Ti's before they ship out.

Certainly interesting but not a large enough dataset to look statistically significant.
yeah ah no.

however i cannot control whatever favoritism you have; people say the same when backblaze reports don't favor their favorite spinning rust - it was a "bad shipment".

pro tip: when the SKU has twice the RMA rate of tens of thousands of units - its not a shipping problem.
 

·
Vermin Supreme 2020
Joined
·
25,774 Posts
2080ti owners more critical to get samsung memory ayy?


return micron since it has an overclocking defect :^)

woah woah woah, hey now. This totally hasn't been done at least hundreds of times.



seriously though, yep it's most definitely not helping those #s.
 

·
ٴٴٴ╲⎝⧹˙͜>˙⧸⎠╱
Joined
·
6,083 Posts
woah woah woah, hey now. This totally hasn't been done at least hundreds of times.



seriously though, yep it's most definitely not helping those #s.
Micron can burn in the ground for this generation




Edit: forgot to mention that super cards generally come with 16gbps GDDR6, which micron I believe have a lot of trouble binning, so it's mostly samsung chips in those




2080ti still using the pisspoor 14gbps speedbin that micron likes to dump its garbage on
 

·
Registered
Joined
·
1,541 Posts
I have to give credit to NVidia for making a 775mm flagship die size. It’s no wonder the RMA rate is so high. I had to RMA one myself. Artifacts on arrival.
 

·
Performance is the bible
Joined
·
7,050 Posts
I have to give credit to NVidia for making a 775mm flagship die size. It’s no wonder the RMA rate is so high. I had to RMA one myself. Artifacts on arrival.
A 775mm die size?
That die size is bigger than a waffer. No wonder you had to RMA it.
 

·
Registered
Joined
·
1,541 Posts

·
ٴٴٴ╲⎝⧹˙͜>˙⧸⎠╱
Joined
·
6,083 Posts
A 775mm die size?
That die size is bigger than a waffer. No wonder you had to RMA it.

wonder what's the yield rate eh?
 
1 - 20 of 59 Posts
Top