Puget data suggests Intel CPUs generally were more reliable than AMD's in 2019-21 (Not Rocketlake!)

MartinPrince · 13 Jan 2022 at 11:40

https://www.pugetsystems.com/labs/articles/Most-Reliable-PC-Hardware-of-2021-2279/

I'm not too sure what to make of it as Rocketlake was very poor compared to 10th gen Intel, Hmm... CPU's have very low failure rates so I probably wouldn't be too worried about it!

jigger · 13 Jan 2022 at 11:51

At a 8-1 offset in retail sales I don’t think we can draw much of a conclusion.

Getting a sense Puget Systems have lot of Xeons and Samsung drives needing to find homes in workstation.

I would be interested to see what HDD’s have the lowest failure rates.

humbug · 13 Jan 2022 at 15:16

MartinPrince said:
https://www.pugetsystems.com/labs/articles/Most-Reliable-PC-Hardware-of-2021-2279/

I'm not too sure what to make of it as Rocketlake was very poor compared to 10th gen Intel, Hmm... CPU's have very low failure rates so I probably wouldn't be too worried about it!

The chart shows a completely different story.

Intel score 8.28 points in failure rates, AMD score 6.87, you would only conclude AMD has a higher failure rate if you ignore 11'th gen entirely, which seems to be what they have done, they seem to be saying "well it seems odd so we are going to ignore it"

11'th Gen Failure rate is near as high as all of AMD's put together.

jigger · 13 Jan 2022 at 15:26

humbug said:
The chart shows a completely different story.

Intel score 8.28 points in failure rates, AMD score 6.87, you would only conclude AMD has a higher failure rate if you ignore 11'th gen entirely, which seems to be what they have done, they seem to be saying "well it seems odd so we are going to ignore it"

11'th Gen Failure rate is near as high as all of AMD's put together.

I get the a sense Puget have over ordered on some kit

tommoT.Striker · 13 Jan 2022 at 18:08

C'mon we can just admit that the highest failure rate was Intel's Rocket Lake.

No surprise Threadripper 3000 fared well.

Hotwired · 13 Jan 2022 at 18:32

That's hilarious that their blanket conclusion on AMD ignores the massive erection of 11th gen on intels side.

In general, as long as we erase 11th gen, amd was worse

Rroff · 13 Jan 2022 at 18:38

You'd hope so from Xeons as they are over-engineered/toleranced and highly tested/binned off the cream of the silicon hence you generally pay a lot for them.

Cyber-Mav · 13 Jan 2022 at 19:45

Just taking the results from this forum alone will show far more threads with amd issues than intel.

humbug · 13 Jan 2022 at 21:20

Cyber-Mav said:
Just taking the results from this forum alone will show far more threads with amd issues than intel.

If people don't buy your crap no one is going to complain about it.

james.miller · 13 Jan 2022 at 21:32

humbug said:
The chart shows a completely different story.

Intel score 8.28 points in failure rates, AMD score 6.87, you would only conclude AMD has a higher failure rate if you ignore 11'th gen entirely, which seems to be what they have done, they seem to be saying "well it seems odd so we are going to ignore it"

11'th Gen Failure rate is near as high as all of AMD's put together.

No, it doesn't.

Puget said:
AMD CPUs in general had higher failure rates than Intel, but we did see an oddly high rate of failures with Intel's consumer-oriented 11th Gen processors... which seems odd, especially next to the very low rates shown by the preceeding 10th Gen.

humbug · 13 Jan 2022 at 21:46

james.miller said:
No, it doesn't.

I'm not blind, AMD's entire range comes to 6.87%, Rocket Lake alone is 6.6% with the rest making up 1.68% for a total of 8.28%.

Either Puget Systems are bad at maths or they are deliberately ignoring Rocket Lake.

james.miller · 13 Jan 2022 at 21:49

humbug said:
I'm not blind

They why accuse them of ignoring 11th gen when they specifically mention 11th gen right under the chart? If you're not blind then read it?

humbug · 13 Jan 2022 at 21:50

james.miller said:
They why accuse them of ignoring 11th gen when they specifically mention 11th gen right under the chart? If you're not blind then read it?

So your contention would be they got the maths wrong?

james.miller · 13 Jan 2022 at 21:52

It's you doing math, not Puget. Puget said AMD were generally worse but 11th gen Intel failures were abnormally high. which is exactly what the chart shows.

humbug · 13 Jan 2022 at 21:53

james.miller said:
It's you doing math, not Puget. Puget said AMD were generally worse but 11th gen Intel failures were abnormally high. which is exactly what the chart shows.

When people trust what is said rather than their own intelligence.

pete910 · 13 Jan 2022 at 21:59

james.miller said:
It's you doing math, not Puget. Puget said AMD were generally worse but 11th gen Intel failures were abnormally high. which is exactly what the chart shows.

Have to agree with humbug on this(Did I just type that! )

In percentages terms that reads as Intel is worse when adding the totals . Something rather striking regards 11th gen being that high .

james.miller · 13 Jan 2022 at 21:59

You're wrong. you made the wrong conclusion. period. Stop trying to sugggest you're more intelligent than people who can actually comprehend what puget said.

james.miller · 13 Jan 2022 at 22:03

pete910 said:
Have to agree with humbug on this(Did I just type that! )

In percentages terms that reads as Intel is worse when adding the totals . Something rather striking regards 11th gen being that high .

This isnt about percentage terms. It's about what Puget did and didn't say (and what Humbug accused them of saying). And you shouldnt come to any conclusion based on any one abnormal data point.

pete910 · 13 Jan 2022 at 22:03

Oh hang on, read it wrong I think, they were referring to in regards the xeon line but the mentioned the shocking 11th gen.

pete910 · 13 Jan 2022 at 22:05

What would have been good if they included AMD's epyc line up to see how that stacked up against the xeon range