High Failure Rate in Radeon 7970s

[sze5003]

The way I purchase cards if Im going to spend a lot of money I just go for a well known brand. Generally ill be weary if more than 40% of the reviews are all bad. Some people rate the card bad because their pc won't turn on because it was bought as a Dell and they bought card x and all of a sudden the card is the main problem.

 

[Makaveli]

Since I'm in the market to get a new high-end Radeon, I've been looking at the offerings on Newegg and what I see has been disturbing. Without going all long-winded, I've seen reported failure rates as high as 50%. XFX is by far the worst offender of DOA and quickly-defective cards. Their "Double D" series of reg and ghz edition 7970s had reported failures at 44 and 50 percent respectively. Gigabyte's GV-R797To-3GD Radeon 7970 Ghz Ed 3GB had 13% of customers report defective cards while ASUS' plain old 7970 saw 39% of customers who wrote a review report a failure.

This is not an exhaustive study. Most of the 7970s I saw for sale (in stock or not) had an average 4-egg review score. And happy customers don't write reviews as often as unhappy ones (it's just human nature). So a study like this doesn't reveal comprehensive truth. It just gives a snapshot.

I almost feel compelled to spend $39 on an extended warranty. Is this just the way things go in graphics cards or are things bad in Radeon land? One thing I would say, at least for Radeon 7000-series cards is stay away from XFX!! They may offer lifetime warranties but that doesn't do you a bit of good if you get multiple cards in a row DOA!

Is all of this is based just on data you got from Newegg ?

That would be like me saying everyone is using 32 bit windows 7 based on the last steam hardware round up.

 

[Chiropteran]

There is a much higher bitcoin mining rate for radeon cards, compared to nvidia.

I wonder if the greater failures (if even true) may just occur because the cards are pushed 24/7 all the time more often.

That said, I've mined hard on 7 5830s, 3 5850s, 1 6970, 1 7970, 1 7790, some for over two years, and zero have failed to date.

 

[Stoneburner]

Some of the posts in this thread are bizarre. What does AMD have to do with the quality of the board? I sincerely doubt there are actual GPU failures. The components (and PCB color) for the boards are the choice of the AIB partners.

 

[Cloudfire777]

^Pretty sure that

This is bad components
and
This is good components

 

[philipma1957]

^Pretty sure that

This is bad components
and
This is good components

since i own 2 of the sapphire and 6 of the xfx that you linked to the xfx have terrible fans and some come with locked voltages. the sapphires are unlocked the fans are great. but all 8 amd chips work well.

xfx should be punished for what they did with these cards. I can use them because I have different coolers and I don't have to under volt them.

Ask me what I think of the 2x sapphires they are great. cool quiet easy to over or underclock.


If you want the xfx and you put this cooler them they work great.

http://www.newegg.com/Product/Produc...82E16835242030

 

[amenx]

Some of the posts in this thread are bizarre. What does AMD have to do with the quality of the board? I sincerely doubt there are actual GPU failures. The components (and PCB color) for the boards are the choice of the AIB partners.

Yep. I believe AMD and Nvidia are roughly equal in quality, its their AIB partners who must shoulder much of the criticism. In that BeHardware survey of returned components, the GTX580's high problem rate was mostly due to one AIB partner, Gainward. As the 7970's problems are mostly XFX. I can recognize that due to intense competition, not AMD vs Nvidia, but rather AIB vs AIB, they may cut corners or user cheaper parts.

 

[Puffnstuff]

Well we should stop to consider that both chips come from the same foundry so if they've got a process issue it will show up in whatever lines are being produced there.

 

[WhoBeDaPlaya]

That said, I've mined hard on 7 5830s, 3 5850s, 1 6970, 1 7970, 1 7790, some for over two years, and zero have failed to date.

I rode my cards hard mining and put them away wet, I mean, on eBay
That said, I never OCed (in fact downclocking the RAM) and kept them well ventilated.

 

[3DVagabond]

if a product works as advertised you are less likely to sit and write the feedback. but on the other hand a person who gets a faulty card or has problems will mostly give a negative feedback to warn others. its human nature to not bother when things work as expected but get very irritated when it does not.

If I owned company A, I would buy a whole bunch of company B's products from Newegg and write bad customer reviews, and a whole bunch of my own stuff and say it was awesome. Just saying.

 

[Will Robinson]

Scott Wasson of Tech Report is considered a trusted source here right?
His opinion of New Egg reviews is that they "have been completely gamed".
That was a year or two ago...

 

[Keysplayr]

Scott Wasson of Tech Report is considered a trusted source here right?
His opinion of New Egg reviews is that they "have been completely gamed".
That was a year or two ago...

Do you trust Scott Wasson?

EDIT: Thought so. Not gonna touch this with a 20ft. cattle prod are you...

 

[Danno21]

A slow and horrible RMA process can really kill a company for me. AMD's seem to have so-so companies making their cards that have difficult RMAs and that's why I stay away from AMD.

I buy Nvidia for EVGA because you speak to people in California and their RMA process is superb. It's a matter of preference, but EVGA has always been there for me when I needed it and it stuck with me.

 

[wand3r3r]

A slow and horrible RMA process can really kill a company for me. AMD's seem to have so-so companies making their cards that have difficult RMAs and that's why I stay away from AMD.

I buy Nvidia for EVGA because you speak to people in California and their RMA process is superb. It's a matter of preference, but EVGA has always been there for me when I needed it and it stuck with me.

Which companies are so-so?

 

[Cloudfire777]

If I owned company A, I would buy a whole bunch of company B's products from Newegg and write bad customer reviews, and a whole bunch of my own stuff and say it was awesome. Just saying.

Scott Wasson of Tech Report is considered a trusted source here right?
His opinion of New Egg reviews is that they "have been completely gamed".
That was a year or two ago...

Yeah right. Go put on the tinfoil hat

 

[Will Robinson]

Do you trust Scott Wasson?

EDIT: Thought so. Not gonna touch this with a 20ft. cattle prod are you...

As a matter of fact I do trust both him and his well run website.:thumbsup:

 

[Will Robinson]

Yeah right. Go put on the tinfoil hat

ugh...you really are a nasty piece of work aren't you....
Now I understand what the mods at Overclock were on about.
Still, you said it best yourself I suppose...:whiste:

Originally Posted by Cloudfire777 View Post
Got banned on overclock.net (some moderators there didnt like me, lol) so I`m moving in here. Hope you don`t mind

 

[Cloudfire777]

ugh...you really are a nasty piece of work aren't you....
Now I understand what the mods at Overclock were on about.
Still, you said it best yourself I suppose...:whiste:

I have no problem with the moderators here. I have a problem with users like you who spread around wild theories that have absolutely no merit in reality.

Companies that buy a bunch of products, verify themselves as owners just to write bad reviews about the product. Do you even understand how crazy that sounds like?
Amazon could be prone to operations like that since you don`t have to verify yourself (but they don`t buy the products), NewEgg on the other hand is safe. They track each purchase from the site and you bet they check history and so does the companies that makes the products.

 

[-Slacker-]

This is not an exhaustive study. Most of the 7970s I saw for sale (in stock or not) had an average 4-egg review score. And happy customers don't write reviews as often as unhappy ones (it's just human nature). So a study like this doesn't reveal comprehensive truth. It just gives a snapshot.

Then don't make "manufacturer x's video cards don't work" threads on a forum that's buzzing with fan boy wars.

 

[blackened23]

I have no problem with the moderators here. I have a problem with users like you who spread around wild theories that have absolutely no merit in reality.

Companies that buy a bunch of products, verify themselves as owners just to write bad reviews about the product. Do you even understand how crazy that sounds like?
Amazon could be prone to operations like that since you don`t have to verify yourself (but they don`t buy the products), NewEgg on the other hand is safe. They track each purchase from the site and you bet they check history and so does the companies that makes the products.

Huh? You don't need to purchase a product on EITHER amazon or newegg to write a review. They're completely the same. If you purchased it, your review will be tagged as a verified buyer. Newegg and Amazon are both the same. Conversely if you want to write a review without a purchase both amazon and newegg, they will let you.

 

[Cloudfire777]

Huh? You don't need to purchase a product on EITHER amazon or newegg to write a review. They're completely the same. If you purchased it, your review will be tagged as a verified buyer. Newegg and Amazon are both the same. Conversely if you want to write a review without a purchase both amazon and newegg, they will let you.

They will remove your ratings after a while if you don`t verify yourself at Newegg. Some might slip through but they do delete them after a while.

 

[blackened23]

Sounds like insider knowledge ^_^

 

[dalewb]

Then don't make "manufacturer x's video cards don't work" threads on a forum that's buzzing with fan boy wars.

I've already explained myself... This was a 3-hour study I did for myself so I could try to get info on which card would be the most reliable within a given card type (AMD 7950, 7970, etc) based on my interests. I came here because I was puzzled by the results I was seeing.

Others have asked why I am doing a study on Newegg user reviews. Who doesn't? Besides the 3 guys on this forum who say they are baseless, I mean. How is anyone to know who creates these posts - whether they are from industry trolls trying to defame their competitors or normal end-users, except where they are verified owners?

Besides, I don't know of any other way to get reports on a products reliability except through consumer reports (they like that sort of stuff but are often way behind on new tech). If the only thing this thread accomplishes is to educate people on where to find accurate reliability data, then I'll consider it a success - that's all I'm after.

 

[Subyman]

I hope the OP knows that 50% failure rates are absurd. Going through and counting the reviews that mention their card died vs how many reviews there are total doesn't say much. You'd have to go straight to the manufacturers or get the RMA processes from Newegg to figure the actual rates. You could do a comparative analysis, but you won't walk away with absolute numbers.

 

[Headfoot]

It's not a study at all. Where is your data?

You got better data? Then show it. Otherwise comment on the data he HAS presented.