Puget Systems Releases CPU Failure Report: AMD CPUs Achieve Higher Failure Rate Than Intel 13th and 14th Generation

[AusWolf]

Four pages so far of the AMD cult losing their sh1t. This thread didn't disappoint.

Considering even the title of the article, and Puget's skewed data with their own power limits, whatever they may be, I'm not surprised. It is designed to put fire under the AMD cult.

I'm sick of this constant "but but Intel" and "but but AMD" tug of war, honestly. We have more than enough of it on the forum even without articles like this.

 

[Why_Me]

Considering even the title of the article, I'm not surprised. It is designed to put fire under the AMD cult.

I'm sick of this constant "but but Intel" and "but but AMD" tug of war, honestly. We have more than enough of it on the forum even without articles like this.

I live for these articles. I go to this site and I scroll down on the 'recent headlines' to find out what thread is going to be off the hook which is usually any thread with Intel, Nvidia and/or Microsoft in the title. Then I guess how many pages it's up to before I click on it to see how close my guess is.

 

[evernessince]

Being this was an article they posted on their site intended to calm their own customers (both past customers and future customers), it makes total sense they'd include that data. I work with people who buy workstations but know nothing about computers. They see headlines though and go "oh, so Intel has problems, right?" and then Puget feels they have to put an article like this out. They (correctly in my opinion) assumed one of the next questions from their customers would be "ok, well what about AMD?". There's still a lot of "AMD? I'm not sure...all my workstations have been Intel" mindshare out there in the business world. Including the data shows their customers that they track it, catch most of the AMD issues in the shop before they get shipped out, and also proves the point they're trying to make that they are managing the intel situation in a manner that keeps failures at a level consistent with other skus.

As you pointed out, people are going to see the headlines 'puget says AMD CPUs fail at 2x the rate'. I can understand wanting to get ahead of an issue but they need to ensure that communications with customers are concise and clear. They simply did not spend enough time explaining the AMD numbers. They mostly left that graph up to the interpretation of the media and they went wild with it.

Now instead of getting ahead of a potential issue they've very likely created concern among their AMD users. Even if their intent was good to begin with, the end result is they've seeded unfounded concern.

 

[Crackong]

AMD CPUs Achieve Higher Failure Rate Than Intel 13th and 14th Generation
Intel 11th Generation failed big time in both performance and relability.

 

[AusWolf]

I live for these articles. I go to this site and I scroll down on the 'recent headlines' to find out what thread is going to be off the hook which is usually any thread with Intel, Nvidia and/or Microsoft in the title. Then I guess how many pages it's up to before I click on it to see how close my guess is.

Yeah, they always get derailed sooner or later. This is the world we live in today, unfortunately. I just don't get why the need to hunt for derailment right in the article itself.

 

[InVasMani]

He’s on the Board of Advisors, which is a representative group of system builders that I guess offer feedback to Intel and Intel partners. That is WAY different than being on the Board of Directors. You generally don’t see BOD’s blogging. Still might present a conflict of interest, but then again, maybe he’s just fulfilling his role by offering guidance to system builders.

https://www.intel.sg/content/dam/www/public/apac/xa/en/documents/brochures/board-advisors-brochure.pdf

Bit of a slip up though doesn't change a whole lot looks bad either way.

 

[chrcoluk]

These numbers are likely derived from a small sample size hence the inconsistencies. Even their "low" failure rate CPU families are about twice the industry average. They aren't useful for anyone but Puget. You could have easily titled this article "Puget shows 11th gen failure issues" and it would have been just as misleading. The reason we got the headline we got is for the clicks, plain and simple.



Intel's official documentation for the 13th and 14th gen actually recommend against the base line profile.

View attachment 357619

This is because if Intel recommended the base line profile as default they would loose a massive amount of performance. The result is that the baseline profile isn't actually the baseline profile and there really is no singular recommendation from Intel in regards to what's safe. (More info on this in GN's latest video on the topic).

Their original article details why they manually go through and validate their own power settings, it's just that it's been taken out of context by tech news websites.



This data isn't likely to be corroborated because it also implicates the 11th gen as having a failure rate of nearly 2 gens combined, which I have seen no reports of. Puget's failure rates in general are twice the industry average. I said it before but these numbers are highly specific to puget and they aren't intended to demonstrate high AMD failure rates.



I very much doubt this hold true after this recent fiasco. 50% failure rate on your gaming servers will absolutely get you fired when the competition is 1.2%.



Correct and the original article wasn't intended to to draw any conclusions on AMD rates. It's the tech Media that is trying to do so for the clicks.

Have you have already decided what is going on, what the failure rates are and so forth?

The pugent data being different to others is explained in the article itself, they dont use stock bios settings.

The failure rate for Raptor lake currently is clearly lower than some of these claims, if it was e.g. 50%, the internet discussions would be wildly different to what they are now, e.g. we have one confirmed TPU user with a failed chip out of what I have seen a dozen or so Intel owners, Jay2cents says his i9 might be potentially affected as it has started becoming unstable, as well as buildzoid (although buildzoid's chip was unstable from day 1, and becomes stable when run at Intel spec, so his isnt in the degradation category for me) but none of the other tech youtubers have reported issues.

Other named sources are gaming companies. Of which we dont know really what their sample sizes are either, whether they too small or not as one of the arguments to discredit puget. Their systems are probably running at bios stock settings or close to it.

On Intel baseline, performance, extreme, even in performance and extreme spec, there is bios's out of spec.

Puget also have shown that a previous gen Intel has highest failure rates in their systems, not a AMD CPU.

Where I do agree is that 14th gen being the newest gen, that the bar might well continue to increase as its new, but this of course will also apply to AMD 7000 series.

Its not hard to find failure reports for any product, but usually such reports dont gain much momentum, they put down to unstable RAM or whatever, of course right now there is massive momentum on these Intel issues, which means suddenly anyone who has "any" kind of instability suddenly has a faulty chip.

 

[Nater]

Considering even the title of the article, and Puget's skewed data with their own power limits, whatever they may be, I'm not surprised. It is designed to put fire under the AMD cult.

I'm sick of this constant "but but Intel" and "but but AMD" tug of war, honestly. We have more than enough of it on the forum even without articles like this.

It's not the first time. They've been roasted before, go look at the comments from their older Ryzen articles. They'd do head to head and neuter the AMD systems with the same 2666Mhz RAM speed and horrible timings as the Intel rigs. Not use PCIe 4.0 vs PCIe 3.0 drives so it's "apples to apples". They had to re-do some of their head to heads because enough of us went "WAIT, that's not how this works". They've always seemed like amateurs to me when it comes to their benchmark reviews, and have learned nothing from people we actually respect the opinions of. (Kyle@HardOCP, Steve@GamersNexus, etc)

 

[Darmok N Jalad]

I live for these articles. I go to this site and I scroll down on the 'recent headlines' to find out what thread is going to be off the hook which is usually any thread with Intel, Nvidia and/or Microsoft in the title. Then I guess how many pages it's up to before I click on it to see how close my guess is.

I'm trying to use my brain on this one, but it is Munday. Data analysis is a big chunk of what I do for a living. The biggest misuse is that correlation does not mean causation, and a really nice representational graph:

Granted, that r squared could stand to be a bit higher before I'd consider it accurate.

 

[Sound_Card]

View attachment 357586



View attachment 357587



Nahh, it's your imagination for sure.

This is basically the end of the thread. Everything else after this post in regard to AMD failure rate is cope, hope, and dope.

 

[solidusKirbo]

Is the failure rate of Intel 11th gen CPU's really that high? Damn, I'm cooked...

 

[A&P211]

Puget systems are just attention whores

 

[LabRat 891]

AFAIK, these stats are on 'OC'd' and tweaked CPUs. Puget is one of the few companies that exists to build Overclocked machines for professional use.
Puget is semi-local to me. Few-years back I had the delusion I might be able to work there, and researched their builds and articles some.

I will say tho... my 1st boxed R5 5600 lost a core; first time ever having a CPU 'go bad'.
IMO, the small-lithography 'highly dense' silicon we see today, implicitly is going to have a higher failure rate (major design-level defects, aside)

 

[FoulOnWhite]

Yeah, they always get derailed sooner or later. This is the TPU we live in today

More realistic. it's ridiculous really, how there cannot be a civilised discussion of either camp without it ending up a squabble and closed thread

 

[AusWolf]

More realistic. it's ridiculous really, how there cannot be a civilised discussion of either camp without it ending up a squabble and closed thread

It's ridiculous how there have to be "camps". I mean, having a preference is fine, but that doesn't make the competition any worse or evil, not to mention it doesn't give you any reason to join a "camp". My preference was Intel in the Pentium I-III era, then AMD with Athlon 64, then Intel again with Core 2 and Core i up to 11th gen, now it's AMD again, which may or may not change again when Intel releases Bartlett Lake.

I've always thought "buy whatever suits your needs". For-profit companies are not your friends. Not Intel, nor AMD, heck, not even Puget Systems. They're all in it for the money, nothing else. Why people can't comprehend this simple fact is beyond me.

 

[sKam]

 

[Crackong]

Why people can't comprehend this simple fact is beyond me.

because 'I am sorry' 'I am wrong' 'I've made a mistake' are the simplest yet most difficult thing to admit.

 

[AusWolf]

because 'I am sorry' 'I am wrong' 'I've made a mistake' are the simplest yet most difficult thing to admit.

Who was wrong, who made a mistake, and who has to be sorry for what? It seems I really don't get it.

 

[1d10t]

I don't think it's necessary to add other members have posted. I just wanted to post an old link from Puget explaining what they mean by Shop Failures from their previous timeframe between 2018 and 2019.

https://www.pugetsystems.com/labs/a...dware-in-our-puget-systems-workstations-1550/

Funny, their report of DOA units seem way below reviews on the outside like NewEgg for example.

https://www.newegg.com/core-i7-9th-gen-intel-core-i7-9700k/p/N82E16819117958#IsFeedbackTab

Just another User Benchmark website, nuff said.


-= edited=-
Forgot to post meme

 

[Vayra86]

Well after this article I say Fugget to Puget Systems.

Keep your drivel where the sun don't shine, and your product too.

 

[AusWolf]

Link - Puget Systems is right, there are no stability issues on 13-14th gen Core. Oh wait...

 

[LittleBro]

I don’t think it was malice, but probably negligence and/or a rushed design. Anytime we see a “refresh” product, it’s typically Plan B from the company. It also usually means it’s not as well-planned or executed, like a rebrand with bumped up clocks or moving a SKU down a level. We’ve seen rebrands from all the big players, and I can’t imagine it was anyone’s first choice.

There are some souces of information or rumors that might shed some light into this ...

 

[yfn_ratchet]

I can't really make heads or tails of what Puget has put out and I think that's because the level of context is about on par with AMD's marketing department. There's so much I don't know about how they reached these numbers, how they configure/tune their systems, what constitutes pass/fail that I have a better time accepting the single sample of that one rant from a guy that burned out a 14900K on a blade running a Minecraft server of all things simply because he provided those criteria.

 

[thesmokingman]

I can't really make heads or tails of what Puget has put out and I think that's because the level of context is about on par with AMD's marketing department. There's so much I don't know about how they reached these numbers, how they configure/tune their systems, what constitutes pass/fail that I have a better time accepting the single sample of that one rant from a guy that burned out a 14900K on a blade running a Minecraft server of all things simply because he provided those criteria.

All you need to know is the CEO of Puget is on Intel's board of directors and he's got a duty to defend Intel, not AMD.

 

[Bobaganoosh]

As you pointed out, people are going to see the headlines 'puget says AMD CPUs fail at 2x the rate'. I can understand wanting to get ahead of an issue but they need to ensure that communications with customers are concise and clear. They simply did not spend enough time explaining the AMD numbers. They mostly left that graph up to the interpretation of the media and they went wild with it.

Now instead of getting ahead of a potential issue they've very likely created concern among their AMD users. Even if their intent was good to begin with, the end result is they've seeded unfounded concern.

That's not Puget's headline. Theirs is "Puget Systems’ Perspective on Intel CPU Instability Issues". They don't mention AMD in the title and most of the article they wrote (for their customers) is about the intel stability issues which they say they have seen. Actual quote: "At Puget Systems, we HAVE seen the issue, but our experience has been much more muted in terms of timeline and failure rate." I'm not even trying to defend them here, but I'm just sick of the BS spin that always happens. What they put out was a marketing piece about how their experience and test methods should put their customers at ease. Whether it would successfully calm current or future customers is up to those people, but the fact that TPU and others have just spun it into "OMG YOU GUYS THEY SAID AMD IS WORSE!" is just misleading. I personally think we're so far into click-bait territory on almost every article out there these days that people need to learn to be smarter about it and either call it out more or just know that most of the article titles you read are likely not representative of the actual facts. What's funny to me here is that Puget didn't even go with click-bait. It's just hand-wavy marketing in hopes to appease their concerned customers. It shouldn't have been taken as more than that.

There's a real conclusion we could draw here (not that it's new), which is that Intel could have avoided a lot of issues on Raptor lake if they made a decision ahead of time to keep the power usage at a reasonable level by default and set up the micro-codes (and default BIOS settings for board partners) accordingly. The data from Puget proves that to be true. It doesn't mean there still wouldn't be issues here and there, but it shows that at least some of the issues are from Intel not putting out any clear rules on default BIOS settings or V/F behavior in their own micro-codes.

It's not the first time. They've been roasted before, go look at the comments from their older Ryzen articles. They'd do head to head and neuter the AMD systems with the same 2666Mhz RAM speed and horrible timings as the Intel rigs. Not use PCIe 4.0 vs PCIe 3.0 drives so it's "apples to apples". They had to re-do some of their head to heads because enough of us went "WAIT, that's not how this works". They've always seemed like amateurs to me when it comes to their benchmark reviews, and have learned nothing from people we actually respect the opinions of. (Kyle@HardOCP, Steve@GamersNexus, etc)

I completely agree with the first part here about Puget. They have been very slow to even entertain the idea of AMD systems and had a skew towards Intel that was only exceeded by the likes of Dell/HP/etc. That said, Steve of Gamer's Nexus is someone people need to be careful of when watching their content. I watch it too and most of the time he's spot on with things, but he gets carried away sometimes and then starts this spiral of nit-pick and blowing things out of proportion when he gets stuck on a topic. The fact that intel has been so wishy-washy with default BIOS settings for a long time has been probably the biggest annoyance of Steve that I've seen and he gets angrier about it every time the topic comes up. It's a real issue that leads to discrepancies in failures, performance, etc. and he's right about that, but he then takes it soooooo far that logic and reason get left behind. We all have our biases and I think his are a little harder to spot, but if you watch his reviews, there's usually something for each architecture, product, whatever that bothers him and it'll skew the whole review. Getting people to watch their videos (and buy their merch) is their business and income, so they do sometimes go into full click-bait titles, thumbnails, and even sometimes the stories are a bit exaggerated at points or over-the-top. I'm not saying he's terrible, but I think people take it as 100% fact or look at it as unbiased every time, when you always need to consider the facts for yourself and draw your own conclusions because even Steve is biased and sometimes wrong (or at least exaggerating) about things because of that.

I've always thought "buy whatever suits your needs". For-profit companies are not your friends. Not Intel, nor AMD, heck, not even Puget Systems. They're all in it for the money, nothing else. Why people can't comprehend this simple fact is beyond me.

^Exactly. That right there is why Puget put their article out. There are all these videos and articles out about Intel having issues and they sell primarily intel machines so they felt they needed to calm their customers down. It's just about protecting the revenue for them. I didn't see anything in their article that was really taking sides or saying "AMD BAD" lol, just trying to claim that even though intel is having issues, they have a "unique" and special approach that makes them the company to give your money to, whether you want an AMD or Intel workstation.