• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

The thread which sometimes talks about RDNA2

Status
Not open for further replies.
Soldato
Joined
17 Aug 2009
Posts
10,719
Like I said, it's a time slice but if people want to make claims using the worst kind of data available from it, there are better supported views which contradict it.

Don't the newest Nvidia cards have the lowest RMA (20 Supers) they're the only cards that have been out same length of time as AMD's 5XXX?

That's the point. They are a refresh and have gained reliability over the first take (2000 non super) which has worse RMA rates than AMDs first (and only so far) take on the 5000 top end.

By "new" I am referring to the initial models produced of a series. The ones that you'd expect to be taking the most risks to get performance.
 
Last edited:
Caporegime
Joined
18 Sep 2009
Posts
30,111
Location
Dormanstown.
Like I said, it's a time slice but if people want to make claims using the worst kind of data available from it, there are better supported views which contradict it.



That's the point. They are a refresh and have gained reliability over the first take which has worse RMA rates than AMDs first (and only so far) take on the 5000 top end.

Well no, it's not that simple.
Nvidia 2080 has been out longer so it's had longer to fail
A 5700XT at the same age as a 2080 in time may have the same or higher percentage failure.

There's also the fact that since the Nvidia GPU:s have been superseded what they're selling has changed and the quantities.

It may be that they sold 50K in X period but 100K in Y period. If the failure's of the last 12 month come from the total pool of GPU's but the failure rate is compared to the 50K figure it's going to look worse.
 
Soldato
Joined
10 Oct 2012
Posts
4,421
Location
Denmark
The easiest way to get better numbers is a larger cooler.
Assume posted before but thoughts?

7JIuJYS.png
It doesn't have to be ugly but they threw that in for free.

I don't mind a lady with some heft to her but this ain't it. "Det er en om'er!"

/Disclaimer: If you are a ********* and feel offended there is nothing I can do for you, sorry.

EDIT: Wow censoring that word is just strange and over the top PC.
 
Associate
Joined
19 Jun 2017
Posts
1,029
Well no, it's not that simple.
Nvidia 2080 has been out longer so it's had longer to fail
A 5700XT at the same age as a 2080 in time may have the same or higher percentage failure.

There's also the fact that since the Nvidia GPU:s have been superseded what they're selling has changed and the quantities.

It may be that they sold 50K in X period but 100K in Y period. If the failure's of the last 12 month come from the total pool of GPU's but the failure rate is compared to the 50K figure it's going to look worse.

Yeah I had done something similar with short span data. You'd need a panel for each card sold - date sold, current state, date of complaint and then fit weibull distribution on it in a Bayesian sense. The industry generally uses mean time to failure as a reportable metric and not RMA rates. This looks like the kind of stats I have come to expect of techtubers :D
 

Deleted member 66701

D

Deleted member 66701


Data is old and doesn't include many 2nnn series cards and we all know the problems those cards had. 5.3% RMA rate for the 2080ti - disappointing for a 1k card. Its also not very good data, as there is no reason listed for returns.
 
Soldato
Joined
17 Aug 2009
Posts
10,719
But you're the one who took the data to make what I'd consider a flawed point about failure rates for the 20XX none supers

It's absolutely flawed, there is incomplete data.

We're looking at a time slice of products at different times in their life, some are refreshed which has implications to its reliability, some are high performance parts, some are not.

Yet it's so easy to make extrapolated claims.
 
Soldato
Joined
18 Oct 2002
Posts
4,333
This video tests different CPU and the AMD 6800xt.

Just interesting but the performance of the 9900k is lower than reviews for the 6800xt with a 9900k.

I wouldn’t trust the differences too much the 9900k is very heat and setup sensitive, it is the best and worst CPU i have ever owned.

The only way we will find out is when people in this forum (and others like it) test it for ourselves and use manually hand tweaked settings.
 
Associate
Joined
21 Oct 2013
Posts
2,061
Location
Ild
Local store says they have no stock of AIB big Navi cards, worse launch than Mvidia once again ETA is mid December for stock :(
based on what? queues are still in the 1000s for ampere 2 months later. But your local coop said they have no big navi then I suppose we can't argue with that.
 
Caporegime
Joined
24 Sep 2008
Posts
38,322
Location
Essex innit!
Tough when your normally the one picking up... jib boom! :D
Terrible lol :D

Anyways, moving on....... Failure rates.... Who cares? I understand that it is frustrating to be without a card for a bit but as long as it gets replaced, it is all good. I am a fan of Zotac and the 5 year warranty. Any from AMD have that?
 

Deleted member 66701

D

Deleted member 66701

Local store says they have no stock of AIB big Navi cards, worse launch than Mvidia once again ETA is mid December for stock :(

Oh look, a post from Grim disparaging an AMD card.

Quelle surprise.
 

Deleted member 66701

D

Deleted member 66701

I don't mind a lady with some heft to her but this ain't it. "Det er en om'er!"

/Disclaimer: If you are a ********* and feel offended there is nothing I can do for you, sorry.

EDIT: Wow censoring that word is just strange and over the top PC.

Well, it is widely considered as an insult and slur, so it's hardly surprising it's censored.
 
Status
Not open for further replies.
Back
Top Bottom