News Intel announces an extra two years of warranty for its chips amid crashing and instability issues — longer warranty applies to 13th- and 14th-Gen C...

[YSCCC]

I cannot speak for anyone else, but since I never run default settings on any CPU, wouldn't you say that I should stick with Intel if I cared about stability?

I cannot speak for you or others, but if it is me I would say NO to stick with intel if you cared about stability and not say, absolute top speed where the crown goes from one to aother gen by gen.

For stability, it should be the vendor who definitely have the say especially they dominated the market that the board partner are mandated to enforce the default auto setting specced by them to be the "default", not any boost or crazy tweak by the board partner team. that is by definitely, by default and should be guaranteed stable out of the box. The K SKUs are meant to be for the tweakers, who are supposed to tweak progressively on top of the knowledge of the original safe values of settings upward at their own risk.

I personally know a lot of ppl who buy the K SKUs and use them at stock at first is because
1) it is the fastest upon release
2) when it gets slow a few generations later, you can OC it and get another a year or two of useful life out of it before upgrading the whole system

This way it will stay stable for say, 3-5 years in default, and get another 1 year of useful overclocked life.

But this time round Intel showed a bit F you for those early 13900k/ks users who trusted them the default should be the default an safe, just to discover issues pop up left and right, being kicked around for months and finally here comes all these "above and beyond RMA".

If only one MB vendor, say Asus, defaulted their bios to pump too much and boost too much to make them look like a better board, it's Asus's fault, if everyone else get the same issue, it's Intel's issue who don't stop them from doing so, if even Intel don't know that the default is problematic, it is even worse. I don't know how they handle those "burnt" CPUs after the recent coverages, but back in early 13900k era, I had a colleague who bought one, getting really unstable for stock setting, send back to RMA, identified to be "overclocked to death, no warranty" and he then just get mad and literally hammered his Z790 and get a set of 7950X instead and never looked back, the 13900k got him unstable computer for 3 months, the 7950X works perfectly fine for a year and rolling, it may be slower than 13900k in certain usecase, but definitely way more stable than RPL

 

[TheHerald]

I cannot speak for you or others, but if it is me I would say NO to stick with intel if you cared about stability and not say, absolute top speed where the crown goes from one to aother gen by gen.

For stability, it should be the vendor who definitely have the say especially they dominated the market that the board partner are mandated to enforce the default auto setting specced by them to be the "default", not any boost or crazy tweak by the board partner team. that is by definitely, by default and should be guaranteed stable out of the box. The K SKUs are meant to be for the tweakers, who are supposed to tweak progressively on top of the knowledge of the original safe values of settings upward at their own risk.

I personally know a lot of ppl who buy the K SKUs and use them at stock at first is because
1) it is the fastest upon release
2) when it gets slow a few generations later, you can OC it and get another a year or two of useful life out of it before upgrading the whole system

This way it will stay stable for say, 3-5 years in default, and get another 1 year of useful overclocked life.

But this time round Intel showed a bit F you for those early 13900k/ks users who trusted them the default should be the default an safe, just to discover issues pop up left and right, being kicked around for months and finally here comes all these "above and beyond RMA".

If only one MB vendor, say Asus, defaulted their bios to pump too much and boost too much to make them look like a better board, it's Asus's fault, if everyone else get the same issue, it's Intel's issue who don't stop them from doing so, if even Intel don't know that the default is problematic, it is even worse. I don't know how they handle those "burnt" CPUs after the recent coverages, but back in early 13900k era, I had a colleague who bought one, getting really unstable for stock setting, send back to RMA, identified to be "overclocked to death, no warranty" and he then just get mad and literally hammered his Z790 and get a set of 7950X instead and never looked back, the 13900k got him unstable computer for 3 months, the 7950X works perfectly fine for a year and rolling, it may be slower than 13900k in certain usecase, but definitely way more stable than RPL

I agree with all of that, of course Intel can put a stop to mobo vendors madness.

But the question still stands, according to Puget's published failure rates, since I'm not running mobo defaults, if I care about stability I should really really really avoid amd, right? Their CPUs fail at 2x or 3x the rate compared to Intel 12/13/14th gen. 2 to 3x. That's huge.

 

[YSCCC]

I agree with all of that, of course Intel can put a stop to mobo vendors madness.

But the question still stands, according to Puget's published failure rates, since I'm not running mobo defaults, if I care about stability I should really really really avoid amd, right? Their CPUs fail at 2x or 3x the rate compared to Intel 12/13/14th gen. 2 to 3x. That's huge.

I don't think so either, based on Puget alone, their graph showed that 11th gen, which is also intel, have a much higher failure rate, and that they didn't specify what the failure rate's base value, so comparing 2% vs 4% can be with a lot of bias, even looking at Ryzen alone, the 5000 series seems have more field deaths vs shop death which means some sort of degradation in the system or minor issues passing their screening vs 7000 series of almost no field deaths of thte 7000 series.

And yet, you don't know what did they tinkered in their custom system and what is their metric to think that is the best. It can be they tune intel more conservatively compared to AMD or so, so that is not relevant data to any users other than Puget themselves. Say to an extreme if I open a company selling systems with my own secret sauce, where I slap a box cooler on every intel and let them run watercooling profiles, and water cool those AMD with conservative profile, then showing intel fails 100% and AMD have 1% failure, it means nothing for Intel vs AMD stability

However, for DIYers what user experienced out there, especially witth this whole unsafe deafult madness, how and how long did it take them to react properly is what matters to non-Puget consumer really consider for stability. As I said be fore, when AMD literally burn a hole in the socket for X3D, I am glad I didn't go for X3D and go for the "more conservative though power hungry 14900k", but look how quick AMD fixed the issue with proper RMA with board partners, and how right before Lunar Lake comes and the RPL is still in burning fire, this makes all the difference on "stability" concerns.

 

[TheHerald]

And yet, you don't know what did they tinkered in their custom system and what is their metric to think that is the best. It can be they tune intel more conservatively compared to AMD or so, so that is not relevant data to any users other than Puget themselves. Say to an extreme if I open a company selling systems with my own secret sauce, where I slap a box cooler on every intel and let them run watercooling profiles, and water cool those AMD with conservative profile, then showing intel fails 100% and AMD have 1% failure, it means nothing for Intel vs AMD stability

Oh I do know. They have it in the same article in their explanation and they also publish it on every one of their reviews. They are basically following intel specifications, 307a etc.

Puget is selling both AMD and Intel mostly with air coolers, their recommendation is the u12a. Arguing that amd fails 3 times more because they are using subpar cooling is ludicrous. You think they are selling pcs actively trying to make them fail?

 

[TheHerald]

but look how quick AMD fixed the issue with proper RMA with board partners,

They haven't fixed the issue, since they didn't issue a recall. There might still be CPUs that degraded during that time and will die just outside warranty, right?

 

[YSCCC]

Dug int

Oh I do know. They have it in the same article in their explanation and they also publish it on every one of their reviews. They are basically following intel specifications, 307a etc.

Puget is selling both AMD and Intel mostly with air coolers, their recommendation is the u12a. Arguing that amd fails 3 times more because they are using subpar cooling is ludicrous. You think they are selling pcs actively trying to make them fail?

I actually just dug into that article and this is part of what they said:

"You can see that in context, the Intel Core 13th and 14th Gen processors do have an elevated failure rate but not at a show-stopper level. The concern for the future reliability of those CPUs is much more the issue at hand, rather than the failure rates we are seeing today" They did note that at the meantime but it could get worse and will update

basically following intel spec and stayed safe is one thing but this only makes intel looks even worse by the reason that this is not the default and not recommended to consumers, so unless you say you got a new CPU will go through the whole intel spec, they are not more stable, if you do both and tinker every bios settings before using, well, maybe you can stick with intel next time and hope it isn't another 11th gen where failure rate is 4 times higher

I didn't suggested that "Arguing that amd fails 3 times more because they are using subpar cooling is ludicrous", what I meant was we don't know how much they've changed/optimized the settings for intel vs how much they have changed/optimized for AMD, once it involve tinkering by the SI, others couldn't tell if the same system is really better, or just they know how to tune one system better than the other, that's it, the cooling "sample" is just to exaggerate what if I do optimize the systems differently could create misleading results.

They haven't fixed the issue, since they didn't issue a recall. There might still be CPUs that degraded during that time and will die just outside warranty, right?

Nope, as said before, I say intel didn't fix the issue by not recalling is not because that RMA alone is not enough, it is the duration of the problem happened in the field and how much warranty is left, also it is the symptom that makes the issue better/worse.

X3D case is more extreme, it is not electron migration, but heat dissipation to a point the silicon melts through, which, the buldging is obvious and cannot be denied, so the RMA is straight forward, while intel's side, it is more related to unseen degradation which shows basically same behaviour as one have OC to death themselves, partly due to it is both pumping too much voltage/current into the CPU, just it is by us or by intel, so they deny RMA for considerable cases, which most recently just get suddenly accepted after ranting in reddit or so.

Offering RMA with 2 years+ left of warranty and no further complain arose since then for 1 year is a good example of when RMA should be working as intended, for a product getting frustrated RMA posts left and right on the intel community forum for 13900k for as long as they exist and kicking the ball around 2 years and just recently claiming they will extend the warranty is not.

 

[TheHerald]

Dug int

I actually just dug into that article and this is part of what they said:

"You can see that in context, the Intel Core 13th and 14th Gen processors do have an elevated failure rate but not at a show-stopper level. The concern for the future reliability of those CPUs is much more the issue at hand, rather than the failure rates we are seeing today" They did note that at the meantime but it could get worse and will update

basically following intel spec and stayed safe is one thing but this only makes intel looks even worse by the reason that this is not the default and not recommended to consumers, so unless you say you got a new CPU will go through the whole intel spec, they are not more stable, if you do both and tinker every bios settings before using, well, maybe you can stick with intel next time and hope it isn't another 11th gen where failure rate is 4 times higher

I didn't suggested that "Arguing that amd fails 3 times more because they are using subpar cooling is ludicrous", what I meant was we don't know how much they've changed/optimized the settings for intel vs how much they have changed/optimized for AMD, once it involve tinkering by the SI, others couldn't tell if the same system is really better, or just they know how to tune one system better than the other, that's it, the cooling "sample" is just to exaggerate what if I do optimize the systems differently could create misleading results.

Oh of course 13th and 14th fail at a higher rate. At a higher rate compared to 12th gen though, not compared to AMD. Amd fails 5x (that's 500%) compared to 12th gen and around 2x times more than 13th / 14th.

Nope, as said before, I say intel didn't fix the issue by not recalling is not because that RMA alone is not enough, it is the duration of the problem happened in the field and how much warranty is left, also it is the symptom that makes the issue better/worse.

X3D case is more extreme, it is not electron migration, but heat dissipation to a point the silicon melts through, which, the buldging is obvious and cannot be denied, so the RMA is straight forward, while intel's side, it is more related to unseen degradation which shows basically same behaviour as one have OC to death themselves, partly due to it is both pumping too much voltage/current into the CPU, just it is by us or by intel, so they deny RMA for considerable cases, which most recently just get suddenly accepted after ranting in reddit or so.

Offering RMA with 2 years+ left of warranty and no further complain arose since then for 1 year is a good example of when RMA should be working as intended, for a product getting frustrated RMA posts left and right on the intel community forum for 13900k for as long as they exist and kicking the ball around 2 years and just recently claiming they will extend the warranty is not.

So are you saying there is no cpu that got mildly damaged due to the elevated vsoc and as a result will die outside of warranty? How do you know that?

 

[stuff and nonesense]

That's what someone would think, but considering both manafacturers only offer a 3 year warranty (well, now intel 5) clearly it's not the case, right?

The manufacture’s warranty covers the purchaser against, so long as it is run within specification, defects resulting from manufacturing processes or defects in materials used in manufacture.
This is the base standard across practically every business.

The time, the lifespan, wear and tear is not normally covered, there may be exceptions, I have not looked.

For any well designed and built device lifespan and failures fit on a curve described as the bathtub. Initially there are a larger number of faults (infant mortality) when the widgets are new (rma, warranty claim). This rapidly falls to a handful* within the widgets expected life, rising again as the widgets get worn out (pay for a repair or replace it, it’s done)

*handful being a very low percentage, those with manufacturing defects fell out during the period of infant mortality. The installed base is solid, stable and barring some unusual occurrence that causes an unforeseen failure, customers will get the expected level of performance from the widget.

Consider a tv, you pay £2000 for a shiny Samsung OLED, warranted for 2 years by Samsung. If it died 730.5 days after the date stamped on your receipt would you buy a second?

A warranty doesn’t imply or infer any lifespan, endurance for a widget.

Question : what do you consider to be a fair and representative lifespan for a cpu, in years?

(Notwithstanding that some people on the bleeding edge will swap out their PC within a shorter periods than the warranty for the flex)

 

[YSCCC]

Oh of course 13th and 14th fail at a higher rate. At a higher rate compared to 12th gen though, not compared to AMD. Amd fails 5x (that's 500%) compared to 12th gen and around 2x times more than 13th / 14th.


So are you saying there is no cpu that got mildly damaged due to the elevated vsoc and as a result will die outside of warranty? How do you know that?

1) all those failure % is well within reasonable range, so 2% for 13/14th gen vs ~4% of Ryzens, the 11th is crazily unstable in that chart but I took with a grain of salt that it is due to some of the Puget's test suite or other bugs for their standard, maybe not at consumer level

2) as said in the now locked post, nobody even intel or AMD could be sure, but said a dozen times before, all the difference is that AMD discovered and "fixed" the issue near the start of product cycle, so it essentially have 2years + of RMA chance before it will be out of production for replacement, plus AM5 can have at least 2 more gens from Ryzen 9000s to replace, so if anything goes bad on one generation, chances are one can fight for a even newer and better CPUs or drop a non X3D and lose some gaming performance for some MT performance, this gives confidence, for RPL, it is near the EOL of the socket, before they will "roll out the true fix", by then those which are degraded to a point of instability, it won't fix, those didn't, have no more option for future LGA1700 new CPUs, just rely on what their stock pile left plus maybe refund, that makes all the difference that the extended warranty is enough or not.

 

[TheHerald]

Question : what do you consider to be a fair and representative lifespan for a cpu, in years?

For a normal office PC / 8 hours a day / basic productivity like office / browsing / databases or something, anything less than 10 years is a failure.

For 24/7 100% CPU usage I really don't know. I'll be amazed if any of the new high end CPUs make it to 5 years - both intel and amd.

 

[ottonis]

It's important to note that a company like Intel would never deliberately make such a generous offer if they weren't sure that this would be the only way to prevent even larger damage down the road. Conversely, this implies that Intel may already be fully aware of the fact that there is a load of defect CPUs to be expected after warranty time runs out, which in turn means that there are lots of invisible damages to CPUs that only wait to become apparent over time.
This is quite disastrous for Intel...so we will see how Intel will do its business in the upcoming few years. Architecturally, they are being outdone by Apple Silicon, by AMD, by nVidia and now even by Qualcomm.. And TSMC is still far ahead interms of production technology.
So, Intel rested on its laurels for the larger part of the past decade, keeping prices high and dismissing meaningful development. Now they have to deal with the consequences...

 

[TheHerald]

1) all those failure % is well within reasonable range, so 2% for 13/14th gen vs ~4% of Ryzens, the 11th is crazily unstable in that chart but I took with a grain of salt that it is due to some of the Puget's test suite or other bugs for their standard, maybe not at consumer level

2) as said in the now locked post, nobody even intel or AMD could be sure, but said a dozen times before, all the difference is that AMD discovered and "fixed" the issue near the start of product cycle, so it essentially have 2years + of RMA chance before it will be out of production for replacement, plus AM5 can have at least 2 more gens from Ryzen 9000s to replace, so if anything goes bad on one generation, chances are one can fight for a even newer and better CPUs or drop a non X3D and lose some gaming performance for some MT performance, this gives confidence, for RPL, it is near the EOL of the socket, before they will "roll out the true fix", by then those which are degraded to a point of instability, it won't fix, those didn't, have no more option for future LGA1700 new CPUs, just rely on what their stock pile left plus maybe refund, that makes all the difference that the extended warranty is enough or not.

Well obviously amd didn't discover and fix the issue that's why both their zen 3 and zen 4 cpus fail at an alarming rate compared to 12,13 and 14th gen.

For some reason GN didn't think it was important to show that graph on his video, he just very quickly skipped over it at the 45 minute mark. Literally - blink and youll miss it,lol. Really makes me wonder why, do you have any ideas?

 

[thestryker]

For some reason GN didn't think it was important to show that graph on his video, he just very quickly skipped over it at the 45 minute mark. Literally - blink and youll miss it,lol. Really makes me wonder why, do you have any ideas?

Steve pretty consistently does that when in rant mode which is why I usually wait until I'm in the mood for those videos despite liking most of their coverage (I have not watched this one yet). I can't think of any times they've lied or omitted stuff, but generally it's downplaying things that aren't in line with the focus. It's very effective for shaming companies, but not the greatest from the consumer side as it feeds the people who don't grasp nuance.

 

[TheHerald]

Steve pretty consistently does that when in rant mode which is why I usually wait until I'm in the mood for those videos despite liking most of their coverage (I have not watched this one yet). I can't think of any times they've lied or omitted stuff, but generally it's downplaying things that aren't in line with the focus. It's very effective for shaming companies, but not the greatest from the consumer side as it feeds the people who don't grasp nuance.

The issue is, according to "popular opinion" intel experiences these issues because "they pushed the chips too far to compete blabla". Now that it turns out amd defaults have a much, much, much higher failure rate than intel defaults, doesn't that mean that amd has pushed their chips far far far further than Intel to compete?

I mean, that's logic 101 right?

 

[stuff and nonesense]

The issue is, according to "popular opinion" intel experiences these issues because "they pushed the chips too far to compete blabla". Now that it turns out amd defaults have a much, much, much higher failure rate than intel defaults, doesn't that mean that amd has pushed their chips far far far further than Intel to compete?

I mean, that's logic 101 right?

What is your source?

 

[Loadedaxe]

Ok, this surely brings me peace of mind, that the vendor is not liable to make sure the board partners enforce their spec as default setting is perfectly fine, since someone with deep knowledge will do their own time consuming tests and find that spec (which was not released with the CPU in any easily accessible manner), and that customers are expected to go into advanced BIOS day one and tinker with like 20+ settings to keep it with low failure rate, thanks to be such a reliable tech company

If you run an AMD you do...

Listen, I get that you really want Intel to "pay".... and your in a loop about their RMA process, but you obviously do not know how business works.

 

[TheHerald]

What is your source?

Puget published failure rates. Zen 3 and zen 4 are leading the charts alongside 11th gen. 10,12 13 and 14th gen have much higher failure rates than the previous 3.

 

[stuff and nonesense]

I cannot speak for anyone else, but since I never run default settings on any CPU, wouldn't you say that I should stick with Intel if I cared about stability?

Nope, I’d suggest that you buy a Puget built pc, one custom built to your specs, tuned by them for stability and never touch the bios.

Reading their site it appears that they trust no-one and assemble pcs with conservative settings allowing the processor to run properly and within spec.

 

[TheHerald]

Nope, I’d suggest that you buy a Puget built pc, one custom built to your specs, tuned by them for stability and never touch the bios.

Reading their site it appears that they trust no-one and assemble pcs with conservative settings allowing the processor to run properly and within spec.

Sure, even then, I should definitely get any intel bar 11th gen over any amd chip, assuming stability is important 😛

 

[stuff and nonesense]

Sure, even then, I should definitely get any intel bar 11th gen over any amd chip, assuming stability is important 😛

Given that both companies have failures even at the conservative Puget settings … the many thousands of chips made that fail from both sides (1% of 30,000,000 rounded values) means there is a whole load of wasted sand…

 

[bit_user]

Puget published failure rates. Zen 3 and zen 4 are leading the charts alongside 11th gen. 10,12 13 and 14th gen have much higher failure rates than the previous 3.

These don't apply to gaming PCs or mass market, in general. They only apply to the professional workstations that Puget builds. Those systems aren't cheap and their customers expect reliability, so it's a good bet that Puget uses workstation boards, other quality components, and tuns them very conservatively.

Sure, even then, I should definitely get any intel bar 11th gen over any amd chip, assuming stability is important 😛

No, because once the system reaches the customer's hands, it's only the field failure rate that matters. So, either go for Intel 12th gen or Ryzen 7000, to achieve the best reliability.

I already explained this several times. You're smart enough to understand the point, which means you're ignoring it and trying to promote an incorrect reading of the data. This is disingenuous and really amounts to posting in bad faith.

 

[Pei-chen]

Unless ASUS also extends my laptop warranty to match, it’s useless to us laptop owners. ASUS BIOS are locked so undervolt is not possible and obviously neither is swapping out a dead CPU

 

[TheHerald]

These don't apply to gaming PCs or mass market, in general. They only apply to the professional workstations that Puget builds. Those systems aren't cheap and their customers expect reliability, so it's a good bet that Puget uses workstation boards, other quality components, and tuns them very conservatively.


No, because once the system reaches the customer's hands, it's only the field failure rate that matters. So, either go for Intel 12th gen or Ryzen 7000, to achieve the best reliability.

I already explained this several times. You're smart enough to understand the point, which means you're ignoring it and trying to promote an incorrect reading of the data. This is disingenuous and really amounts to posting in bad faith.

I said from the beginning, pretty clearly, that if I used intels default (the same way puget does, they even list their settings in their CPU reviews btw, you can basically copy them) I should avoid 11th gen and everything amd like the plague, and go for 12 13 14th if stability is important. What part do you disagree with specifically?

 

[drjohnnyfever]

I hate to ask a novice-type question, but what does this mean for consumers using the i5 chips like the 14600K? I've read about the MB manufacturers setting the limits too high out of the box, but still have some difficulty determining if this is an Intel problem or MB issue.

Is Intel stating that their chips are deficient or that, given the settings of MBs, they are likely to become damaged.

My only experience with this is the entry level gaming PC I built for my son using the i5-14600K and an MSI Pro Z-790A WiFi MB. When posting the first time, the MB wanted the cooling mechanism and then set the Bios based upon that response. I chose the basic cooler, but wound up using a box cooler. I did go back in later, after updating the Bios, to lower the voltages to the chip based on recommendations from users here to address the MB default settings.

I guess the stupid question is: what should an average user, who either doesn't OC or does it minimally, do?

 

[TheHerald]

I hate to ask a novice-type question, but what does this mean for consumers using the i5 chips like the 14600K? I've read about the MB manufacturers setting the limits too high out of the box, but still have some difficulty determining if this is an Intel problem or MB issue.

Is Intel stating that their chips are deficient or that, given the settings of MBs, they are likely to become damaged.

My only experience with this is the entry level gaming PC I built for my son using the i5-14600K and an MSI Pro Z-790A WiFi MB. When posting the first time, the MB wanted the cooling mechanism and then set the Bios based upon that response. I chose the basic cooler, but wound up using a box cooler. I did go back in later, after updating the Bios, to lower the voltages to the chip based on recommendations from users here to address the MB default settings.

I guess the stupid question is: what should an average user, who either doesn't OC or does it minimally, do?

If you choose basic cooler option and stick it most likely you are fine - cause it sets something like a 125w power limit.