The simplistic boost behaviour present on older CPUs is Turbo Boost 2.0, so if you want to disable the more 'excessive' boosting then you can turn off TB 3.0, adaptive boost and TVB, but so far as I know, there's no confirmation anywhere that single core boost or voltage spikes are the cause of the problem. I suspect if you have a broken CPU it'll probably show eventually, no matter what you do.
-
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.
14th Gen "Raptor Lake Refresh"
- Thread starter Dave2150
- Start date
Still early days as time goes on we'll know for sure what's happening, and now if I was getting error it would make me second think
Intel CPUs are oxidising/rusting due to contamination during the cpu production?
Sounds bad
But a massive problem for Intel as Steve said is the coming Zen5 reviews. It's likely that reviewers are going to be reviewing zen5 against 14th gen with 14th gen running on Intel baseline - so Intel's gaming performance is going to get decimated in benchmarks next to zen5
Sounds bad
But a massive problem for Intel as Steve said is the coming Zen5 reviews. It's likely that reviewers are going to be reviewing zen5 against 14th gen with 14th gen running on Intel baseline - so Intel's gaming performance is going to get decimated in benchmarks next to zen5
Last edited:
running at 'baseline' seems better to me than not running at all.
Associate
And the Zen 5 X3D parts are yet to come, so depending on the game whatever Zen 5 will be against Intels baseline you can most likely add another roughly 5 to 15% more performance on top of what Zen 5 will be, again depends on the game so could be less or more. Bad time for Intel as they've lost it just works reliability and will be decimated in gaming, reliability and power consumption.
Last edited:
Soldato
It's 100% unverified, but a plausible theory [oxidization], due to the rate and inconsistency at which the issue occurs, which would be attributed to the rate of oxidation and breakdown. The memory theory is the least plausible, as have seen dozens of cases if not more where the user isn't stable using a single memory module at 4800MT. The fabrication theory is a real issue for Intel, there's no way of validating how many units would be affected.
Last edited:
Though it may be more than one, possibly related, issue, thing is though some systems are throwing up an out of memory error - not necessarily under CPU load, or failing certain stress tests, straight out the box, while others appear to over time degrade to where that starts happening. To me that doesn't seem consistent with an oxidation problem.
Soldato
Depends on where on the die it's happening, doesn't it. There will be multiple red herrings here, too.
If a CPU isn't receiving enough voltage for a given frequency based on the fused VID table right out of the box, which has sometimes been the case, then the most direct question is "why isn't it enough".
If a CPU isn't receiving enough voltage for a given frequency based on the fused VID table right out of the box, which has sometimes been the case, then the most direct question is "why isn't it enough".
Last edited:
Associate
I've been a bit lost with what is going on with these cpus should I be worried as I have two 13900k systems? Should I just offload them while I don't have issues?
Caporegime
Impossible to know because Intel are silent on this.
If its not misbehaving you can't RMA it, if you don't trust it the only option is to sell it and buy a Ryzen CPU, but if that's your thinking why not just wait to see if it does fail? Ok the obvious answer to that is because then the CPU is worthless, I'm trying to think about what i would do if i was in your shoes, i would probably hang on to it and hope it doesn't fail, if it does i'll RMA it.
I will say this, if i was in the market for a new CPU i wouldn't be buying any 13'th or 14'th gen CPU.
Hard to say right now based on the inconsistent information out there so far. I'm yet to find anyone who is seeing this issue widespread like some sources are claiming including a couple of people I know who manage 100s of 13th and 14th gen systems for work (they are seeing more like 3% failure rate) and I haven't encountered theses failures yet on any 13th or 14th gen system I've had anything to do with.
Last edited:
Caporegime
To say my subjective experience is this is fine, nothing to see here is not helpful, you are literally the only one on the internet trying to gaslight people, everyone else agrees there is a serious issue here.
is it something to do with the 'big / little' design?
Just isn't enough information, but the 12th gen have E cores as well without apparently being affected.
that's true, but if you seperate the P cores does this still happen?
I dunno at a purely hardware level, supposedly fully disabling E cores doesn't necessarily fix the issue on affected CPUs - as before not managed to find one myself yet to play with and see what is what.
Wendell said that he 'fixed' some CPUs by disabling the E-Cores, but that there was no consistent way of fixing any of them, so my guess is that a broken CPU is just broken and it is not related to the design.
Associate
The scenario outlined by GN's source, which has Intel produced contaminated CPUs over a certain period is one I find persuasive. That would explain the lack of any definitive fault and 'fix' procedure, because contamination will, by its very nature, result in unpredictable behaviour from the affected parts. And it would also result in some early and more recent chips, manufactured outside the contamination period, not being affected at all and functioning normally.
So, yes, if that's the reality of the situation then broken CPUs are just broken. There's no way (outside of date/batch codes) to know if a chip is affected, how long it will last, what the failure symptoms will be, etc. And any 'fix' of disabling cores, reducing memory speed, etc, is just delaying the inevitable failure. If oxidisation is the root cause the failure analysis commissioned by GN should show it quite clearly - putting my evil capitalist head on, if I were AMD I would quietly offer to underwrite any expenses GN runs up doing the failure analysis. Conclusively proving a manufacturing defect will be gold dust to anyone considering legal action against Intel or a system vendor for selling - knowingly selling - faulty hardware.
Yeah, I agree.
It definitely makes the most sense and it would explain why some CPUs are apparently entirely unaffected.
Hopefully GN can do some data collection with the affected parts and if it turns out they're all between certain dates that will be a slam dunk!
My only concern is what Rroff has mentioned earlier, that they're likely to be a high number of false positives, because there are consistent issues with enabling XMP and high boost voltages/power draw that I believe are outside of the scope of the supposed manufacturing defect.