Intel has a Pretty Big Problem..

[Rroff]

Of course the whole system power isn't dramatically different, it's the whole system. And how about higher external heat dissipation requirement? Plus dissipate from case.

And the PSU angle is clutching at straws.

The whole system power is ultimately what is going to matter - whether CPU or what is using it 99% of that energy is going to be turned into heat.

The PSU angle is largely clutching at straws except in the case of the 14900 where even with conservative power limits it can use dramatic amounts of power, in reality if you look around forums, etc. in most cases people are still speccing up the same kind of wattage PSU at a given performance tier whether buying an Intel or AMD system anyway so it is largely an irrelevant point.

 

[Rroff]

And that's the way Intel likes it, the vagueness means no recalls, no rma

If they are dying like flies there is only so long Intel can hide it.

 

[humbug]

The whole system power is ultimately what is going to matter - whether CPU or what is using it 99% of that energy is going to be turned into heat.

The PSU angle is largely clutching at straws except in the case of the 14900 where even with conservative power limits it can use dramatic amounts of power, in reality if you look around forums, etc. in most cases people are still speccing up the same kind of wattage PSU at a given performance tier whether buying an Intel or AMD system anyway so it is largely an irrelevant point.

The 14900K pulls way more power than my upper mid range GPU, my CPU pulls less than half at worst, if i swap my CPU for an Intel one i will have to upgrade my PSU.

 

[humbug]

If they are dying like flies there is only so long Intel can hide it.

Tell that to the people who are being refused RMA's for degraded CPU's.

 

[jigger]

The 14900K pulls way more power than my upper mid range GPU, my CPU pulls less than half at worst, if i swap my CPU for an Intel one i will have to upgrade my PSU.

Its the same across whole line up. To build a system around Intel requires more attending to and greater expense than AMD. This drives OEM sales.

 

[mattyfez]

Well this thread has gone wildly off topic lol!

 

[malachi]

So who is going to send in their CPU for RMA?

Might doing mine after the August patch.

 

[Freddie1980]

So who is going to send in their CPU for RMA?

Might doing mine after the August patch.

From what Steve was saying on the recent Gamers Nexus video on the matter Intel said it would replace faulty chips but you may have to submit several RMA requests to get it done.

 

[Rroff]

So who is going to send in their CPU for RMA?

Might doing mine after the August patch.

So far had no problems with my 14700K, had it since launch running slightly increased turbo though I've reset that for now to see how this issue shakes out. Don't see any need to RMA as things stand.

Personally yet to encounter this issue at anything like the reports from certain companies - no one in my circle of acquaintances has this problem yet at an individual level, those I've talked to who manage large numbers of these systems (including people who do heavy weight Unreal Engine 5 based production) have only encountered it in a tiny number of systems.

 

[McManicMan]

No problems here, have updated bios to latest and using intel profiles where as before was using ASUS settings. Not noticed any diffidence in gaming or desktop usage, but there is noticeable cinebench drop but not so much in 3dMark. Personally at present not bothered and still have a 2yr warrenty with OCUK anyway

 

[Daytrader]

Yeh, even a nerfed 13900K/14900K by like 10% will only be like a 1fps loss in a game, its mainly the gpu that counts for gaming.

 

[Jay85]

until my 13th and 14th gen chips die, i'm taking everything with a huge grain of salt.

From my look on this so far, it's been those who run stock or enable chip enhancement functions AI overclocking that is killing their cpu's. The people who i've spoke to who tweak and manually overclock or undervolt their cpu's seem to be fine so far.

Afterall we know what core enhancement does, it just throws silly voltages at the cpu which to no surprise will cause degradation over time.

 

[TrixP10]

until my 13th and 14th gen chips die, i'm taking everything with a huge grain of salt.

From my look on this so far, it's been those who run stock or enable chip enhancement functions AI overclocking that is killing their cpu's. The people who i've spoke to who tweak and manually overclock or undervolt their cpu's seem to be fine so far.

Afterall we know what core enhancement does, it just throws silly voltages at the cpu which to no surprise will cause degradation over time.

that must come as a shock to Intel

 

[G J]

Well this thread has gone wildly off topic lol!

Some people are absolutly loving this drama esp the Intel haters and having to sift through all the trolling and off topic to find any actual information/ discussion about this topic is tiresome so I had taken a break from this subforum from Friday until today and skimmed though what I've missed and it's still full of the same ****.

The mainstream tech tube space amongst some truth is full of fear mongering while pretending to care about consumers.... yeah they do but its below farming drama for clicks and money even Gamer's Nexus who put out a great video had to spoil it with his interpertaion of events so half of his video is whatever though his speak with me Intel or I'm cutting you off reaks of jilted ex lover vibes. His ego and inflated self importance is off the charts.

 

[jigger]

Yeh, even a nerfed 13900K/14900K by like 10% will only be like a 1fps loss in a game, its mainly the gpu that counts for gaming.

Back of the envelope calculations but, with the assumed revised settings it’s probably going to be 5-10% depending on graphics card brand, chip quality and game. Obviously the high performance chips will take a bigger hit in the more demanding scenarios and certain system configurations will an effect.

 

[Daytrader]

Ah, i thought that was cpu score loss of 5%-10%, when i had my 6700K and 2800 super and was looking at upgrading to like a 12900K, even that was only a few fps difference at 4K i play at, i would not of thought a couple of 100mhz drop on a cpu would have hardly any impact myself, but i guess benchmarks will be out soon enough for games.

I might add, i upgraded to a 4090 fe also.

 

[jigger]

Ah, i thought that was cpu score loss of 5%-10%, when i had my 6700K and 2800 super and was looking at upgrading to like a 12900K, even that was only a few fps difference at 4K i play at, i would not of thought a couple of 100mhz drop on a cpu would have hardly any impact myself, but i guess benchmarks will be out soon enough for games.

I might add, i upgraded to a 4090 fe also.

TBH we need to wait until Intel decides what is included in the update. We could see a power reduction and completely different boost algorithm.

 

[mattyfez]

TBH we need to wait until decides what is included in the update. We could see a power reduction and completely different boost algorithm.

Maybe it will, probably will, even...

When I first got my 13600k i tried an AI overclock, just for a few bench runs, to see if it was worth fine tuning a proper overclock, plots twist, it wasn't, extra power draw and heat for basically no real world performance gain.

So I went with stock speed & negative offset undervolt... plot twist, no real world performance loss. **BUT** Peak temps are down 10c from stock and peak power draw was down about 15watts. I've not seen my voltage ever go above 1.27 peak running like this, so I'm not really concerned until a few new bios revisions have come out, I'm certainly in no rush to update to an 'interim' BIOS right now, as my system isn't showing any alarming statistics and is totally stable.

 

[jigger]

Maybe it will, probably will, even...

When I first got my 13600k i tried an AI overclock, just for a few bench runs, to see if it was worth fine tuning a proper overclock, plots twist, it wasn't, extra power draw and heat for basically no real world performance gain.

So I went with stock speed & negative offset undervolt... plot twist, no real world performance loss. **BUT** Peak temps are down 10c from stock and peak power draw was down about 15watts. I've not seen my voltage ever go above 1.27 peak running like this, so I'm not really concerned until a few new bios revisions have come out, I'm certainly in no rush to update to an 'interim' BIOS right now, as my system isn't showing any alarming statistics and is totally stable.

It’s difficult to be sure just how of an impact lowering power use will have and qualify a figure because it’s going to depend on the workload and system configuration. Every user will be impacted differently and Intel’s fix has to work across a product stake.

It’s a wait and see situation and power seems to be just one facet of the problem.

 

[Rroff]

until my 13th and 14th gen chips die, i'm taking everything with a huge grain of salt.

From my look on this so far, it's been those who run stock or enable chip enhancement functions AI overclocking that is killing their cpu's. The people who i've spoke to who tweak and manually overclock or undervolt their cpu's seem to be fine so far.

Afterall we know what core enhancement does, it just throws silly voltages at the cpu which to no surprise will cause degradation over time.

It is a weird one, there are 1-2 companies running a lot of 13th and 14th gen chips on W680 boards without MCE, etc. who are reporting very high failure rates but they are not entirely unique in what they are doing and there are loads of other companies doing similar stuff who've not said anything, which they almost certainly would have if they were experiencing failures of that magnitude, and others where someone has commented but only seeing a small number of failures.

My instincts are it is individual CPUs where one or more cores are regularly hitting over 1.5v improperly for the current frequency and thermal situation (doesn't necessarily mean high temperatures), likely due to a faulty algorithm, which are the ones which are failing, possibly there was a bad batch of those but it doesn't cleanly cover the situation above unless there are pools of tray CPUs affected and a fragmented number of retail. Also a total unknown at this point how far it goes and/or if every chip is affected and will degrade prematurely or not.

There also seems to be 3 or 4 issues involved - aside from a supposedly limited batch of chips with the via oxidation problem, there seems to be some older chips with a voltage linked instability problem out the box, which Intel tried to address with a microcode update which then seems to have exposed this issue with a different set of chips.