Summary:https://www.youtube.com/watch?v=H8LrwI-I_fYDeep dive:https://www.youtube.com/watch?v=QzHcrbT5D_Yhttps://www.youtube.com/watch?v=oAE4NWoyMZkTLDR:Don't buy Intel CPUs
>>101419538>Crashes while playing gamesGood. Get real work done
>>101419538Fuck!I only buy prebuilts and last one I got is that KF version, what do
>>101419563>Prime95>Handbrake (video encoding)>Fucking Visual StudioSure...only games.
>>101419583>Prime95>Handbrake (video encoding)>Fucking Visual StudioWow its even better than I thought. Get real work done
>>101419563>OY VEY JUST (((GET WORK DONE))) FOR ME, GOY!
>>101419639>$1000+$1500>=$5500>*saves to disk*Oh boy I'm glad I got work done at SIX GIGAHERTZ!!!
>>101419563 ^ / \: )
so current-heads suffer and waste time watching those movies, what a discovery
You have ONE fucking job Intel. How are you this bad at everything?
>>101419538gaymers deserve to suffer
>>101420219See >>101419583
>>101419538This is what I was saying three months ago on this board and everyone was bashing me for it. Intel's 13th, 14th and probably 15th gen CPUs are degrading, faggots !
>>101420292Even better as benchmark autists should suffer double. when did Intcel become this based?
>>101420337>Video encoding>Programming>benchmarksYou ok, retard?
>>101420345be careful running your dickbeaters too fast over your cheeto encrusted keyboard or you'll trigger another crash hahahahahaha
>>101420617Good thing I don't have an Intel CPU :^)
>>101420312Good thing intel fucked up the next generations naming scheme to make it easier to forget this fuckup and intel completely.
>>101419681>tfw I don't know how to :^)forcepepehands
i just have to pair my $1000 cpu with a bottom end gaymer motherboard that can't supply the rediculous power required for the ultra hyper thread boost gaming mode 7 GHz ok sweetie
>>101419538Core 2 Duo doesn't have this problem.
>>101421929It's probably voltage and not power. That hyper mode is only for a couple of cores. Top end gaymer motherboard with manually configured all-core OC that ends up disabling that feature might work out more stable actually.
>>101419538fanboys get what they fucking deserve: diarrhea all over their face from the company they shill fori still remember the intel shill here claiming everything is fine with his 14900K 4090 rig and everyone else warning him of this problem was retardedwhat a dumbass his cpu is gonna fry and so will his gputo all the goyim that fell for his bullshit refund your intel shit if you can and wait for AMDs ryzen 9000 series release to build a new more energy efficient, stable and faster PCthe 9000 series release is gonna make the 7000 series cheaperyou might snag a 7800X3D (14900K performance at half the power consumption and price) for a good deal
>>101422103>It's probably voltage and not powerit's voltage and current, which works out to 60 W on a single corethat's a lot for such a tiny area
>>101419538Having so many backdoors built in must use a lot of resources.
>>101419563>fun is bad, you have to spend your life toiling away for Mr. Nosenstein
>>101419583getting real work done means joining the marines and shooting boats
>>101422149The problem still occurs on server grade W680 motherboards that use fully stock voltages and underwatt the CPUs to reduce power usage. It's not voltage, current, or power.
>>101422457>that use fully stock voltagesyes, those are the problem>underwatt the CPUs to reduce power usagesingle core boost won't hit any power limits
>>101419639dumbfuck baiter
>>101422140There not issuing refunds or rma's
>>101419583also compilers...
Sucks how it's not safe to be an early adopter of almost any new tech anymore.I know someone's gonna try to look worldly-wise by saying "t'was always thus", but nah. Early adopters often had to put up with some jank, sure, but there wasn't any serious risk that the product would simply be nonfunctional. That seems to be common now.
>>101422656To be fair, what would they do, replace it with another faulty CPU? They're basically up shit creek without a paddle, there's no winning option left on the field for them. They can't replace them with anything, and the defect rate is so high they'd have to do a full recall which would be enormously expensive and a PR nightmare. So instead they're going to stick their head in the sand and pretend there's no problem until the lawsuits start pouring in.
>>101422927>To be fair, what would they do, replace it with another faulty CPU?well, they could always send you a 7950X
>>101422955But then you'd need a new mobo, so you still aren't made whole. This applies even if they send you a Lunar Lake CPU once those come out.
>>101422974then it's time to kneel to AMD's superiority and port zen 4 to LGA1700
>>101422290You mean being shot at on boats.
>>101421929>can't supply the rediculous power requiredOne of the OP videos mentions game servers, which do not overclock and use very standard motherboards, facing the exact same problem.
>>101419538Why are people blaming intel for this one? We only had to implement Thermal Velocity Boost because Zen 4 was too fast. Clearly AMDs fault.
it just works, i had it since late 2015 running at 4.7 ghz. stable innovation brings stable working products
>>101419538Don't care, still going to use my 8th gen Thinkpad because nothing better has come along.
>>101419538Does this affect laptops processors too?
>>101423363No, only high end (high voltage) server and desktop SKUs
>>10142336313900H is allegedly affectedrule of thumb, if it's raptor lake and it runs at high frequencies, it's going to kill itself eventually
>>101423363probably not. laptops only boost to 157W while desktops go to 350W in stock settings
>>101423376>>101423383Ah phew, I bought my younger brother an i7 laptop for uni recently. Hope it doesn’t die
>>101419538INTELAVIVAHAHAHAHAHAHAHAHAHAHAHAHAHAHHAA
>>101419538This post is antisemitic
>>10141953813900KS at 6.1Ghz here for over 1.5 years.Given all the available data + some enthusiast groups with zero problems (all core overclocks day 1).It's safe to assume that the single-core boost pushing 1.5V+ is the culprit. Intel can't do anything about it because they advertised said boost clocks, lmao.People are weird and keep trying to dunk on me. Imagine being a brand loyalist.
>>101419538My pc is fine
>>101423713Yeah like a brand loyalist and Intel shill like you would say the truth for once in your life lol
>>101423746What is brand loyalist about my post?AMD nigger lmao.
>>101419902They had to overclock the shit out of them to beat AMD and now they are paying for it.
Apple won again.
AMD won. ARM won. Crapple won. Snapdragon won.
>>101421929
>>101419583>>101419639Visual Studio is not real work. Real work is done with your hands and buckets of sweat. Or Excel spreadsheets.
>>101423309Haswell was actually one of the rare few modern intel uarchs designed in the USA
>>101423713Yup intel played themselves this time
>>101423842The AMD nigger lmao. See? I knew all Intel loyalists are liars.
>>101424030Yes, you are an brand loyalist AMD nigger.
>>101424080But I am not a liar. A rat. A roach. That's the difference between you a snake and me an AMD Chad and Lisa Su appreciator and Jim Keller worshiper.
>>101424190delusional retard lmaoi just buy whatevers the best at the time every 2 years
>>101423964lelI find it peculiar we have gone from common understanding that CPU's were the least likely component to fail in a PC to this shitshow where both major vendors have exploding and degrading parts.
>>101419563>>101419639Shalom.
>>101424286Which is funny because the best for the past decade has been AMD.
STOP PLAYING GAMES YOU RETARD GAMES ARE FOR CHILDREN THAT YOU ARE NOT
>>101424444Moore's law is dead and silicon limits have been reached despite new FET tech being developed to continue using cheap silicon and not pay genuine gorillions of dollars to switch entire industry + RnD + supply chain to another more suitable material.No, Meme-V or other RISC architectures like ARM wont save you, only delay the inevitable and also give you shit perf while at it than bloated x86.
This thread reeks of antisemitism
>>101419538Wait, people still buy Intel? LOL
>>101424492hey >>101424286
I'm buying one now for the class action lawsuit
>>101424697Last class action lawsuit was the 970 3.5 and Nvidia just gave people like 30 bucks lol
I just bought a bunch of Intel shares thinking they are of the cusp of fucking delivering, bros...I don't even like these retards and I have a Ryzen. Holy fuck, man
>>101425357>inAaaaa I'm going to kms
>>101422290Who the hell wants to be a zogbot in this day and age?
>>101425357that on the cusp of delivering is about the not yet in production fab nodethe things breaking rn and the extremely extended 14nm ++++++ cpus that should never exist to begin with
>>101425419Yeah, but it doesn't bode well for the immediate future.
>>101425447think of it this wayif intel cpus were ever gonna catastropically fail it would be the old zombie node cpus pushed to the limits to compete with amd cpus using tsmc just before the new competitive node
>>101425471Are you neck deep into them too? May the Lord have mercy of us
>>101425573Intel has actually managed to stay admirably competitive with the limitations of their fab failings.AMD has such a huge advantage using tsmc yet people still keep buying intel for some applications, some even buy it for their gaming rigs despite the 3x watt use. When the intel cpu division gets to use an actually good node they'll rape amd straight up.
>>101425596I'll print that and read everyday before I sleep from now on. Shalom, you nosey bastard.
>>101425471Intel 20A only competes with N3E on performance while its density is trash and power is worse
>>101423713Server boards boost to 1.5v+? cpus on them have the highest death rate
>>101425618its not just a transistor density increasebackside power delivery and the gate all around are gonna impact things too and it's interesting to see if 20A does it better than samsungs or tsmc
>>101425628I don't know. Wendell didn't give any info other than that undervolted + power limited.I've just tried it on my consumer MSI board and at 65W it still boosts to 6Ghz on a single core.
>>101419538If this happens only with UE5 games, why isn't it considered a UE5 bug?
>>101419538INTEL'S FINISHED
>>101419538SHUT UP SHUT UP SHUT UPPPPPPPPPPPPPPPPPPPPPSTUPID ANTI SEMITE GOY. STUPID FUCKING CHUD SHUT THE FUCK UP INTEL NEVER CRASHES!!!! I'M CALLING THE MOSSAD!!!!!!!!!!!!!!
>>101425807>If
>AMD breaks records >Intel breaks
I've had my 13700K for a year. Undervolted -0.085v adaptive, adjusted LLC, retarded OC features disabled. Load VID varies between 1.22v (all-core) and 1.3v (peak single-core, usually ~1.28v). Replaced the ILM so it's not bent by the stupid socket either. Air-cooled with a PS120, max ~85ºC in the summer, max package power 214W. No throttling, only limit is V/F. Scores just under 31K in R23.No instability, no signs of degradation that I can tell. I hope that doesn't change, but so far I've been happy with it.
>>101424487>everyone's a jewwwwwHave sex incel.
>101426275the (((jokes))) write themselves
>>101426275>>>101424487>>everyone's a jewwwww>Have sex incel.
>>101419538uninstall the spyware launcher and play a real game then it will not crash.
>>101425992Okay snowflake
>>101419538oy vey why is this antisemitism allowed?
>refund status: 404
>>101419538At this point it's only natural for such a complete self-destruction to end up with AMD somehow managing to self-sabotage an already secured win.
>>101426649lol
>>101426649It's AMD, I wouldn't put it past them.
>>1014266499000 series leaks are showing fantastic gains in temps and performance so unless intelaviv orchestrates another cohencidence AMD is fine for this generation at least
>>101426169I bet the bending ilm is the cause of the "bugs". I would bet my house on it actually.
>>101426275
>>101426671Did you see the latest AMD GPU leaks? There was a thread here a few days ago showing it being like 5x faster than a 4090 in calculations per second.... If it's anything to go by, both Intel and Nvidia have something to worry about.
>>101426685XI'm sure we would have heard something more concrete or more leaks if that was to be real
>>101426699Time will surely tell. That's why I didn't bother to save the image. One can dream. I'd love to see Nvidia shit their pants.
>>101426685>5x faster than a 4090 in calculations per secondyeah noprobably just some dork editing the image to rile people up
>>101426672I was thinking about this too. worth asking on enthusiast forums whether people with contact frames are seeing the same issues
>>101420312If you haven't realized we have a bunch of resident intel copers on here by now, you deserve to be bashed
>>101419639I am also thinking about getting a 14900K for my Excel workload.
>>101426685>leaksholy fuck what year is it? are people still falling for the memes?
>>101426649>9600x $350>9700x $500>9900x $700>9950x $1000Let's go
>>101426685That thread was about performance measurements of the datacentre MI300X beating nvidia's 2023 H100. Not a leak.But its not surprising considering the MI300X uses >3x more silicon than the H100
>>101419538Back to 2500K hell yeah!
>>101419538Sounds like a winfag Fortnite zoomer issue. Not my problem.
>>101426685kekI remember the Zen 2, 3, and 4 "leaks", 90% of which turned out to be fanfiction.
>>101428043Im still waiting for that $200 5ghz 8-core ryzen...
Why would someone need more than a 13600k for gaming with an intel cpu, /g/?
>>101427441Don't forget the motherboards; they'll have a Gen5 link to the chipset. Maybe both of them. Finally reaching parity with Intel PCH bandwidth isn't cheap. :^)
>>101428082480hz
>>101419538Intel shills getting BTFO. You love to see it.
>>101427986>(((stallman)))
>>101419538How is this not a software problem?
>>101423713>>101425745datacenters are not using consumer boards but W680M which are very conservative on voltage and you can't even turn on XMP on itand after all that they still experience multiple failures
>>101429003Because it's a problem with the cpu itself.Now that problem may be firmware, but people more knowledgeable than me have been saying it's likely that they can't fix it with a firmware update.
>>101429003The decompression software used in UE5 crashes a lot but 96.9% of these reports (filtered to desktop cpus) come from i7 & i9 13/14th gen cpusThe last 3.1% includes all other cpu models from amd and intel.
>>101429951so say, a 14th gen i5 should more or less be in the clear?
this will hopefully be the end of this big little trash
>>101423309I actually had one of these degrade and then later die entirely due to a Gigashyte motherboard. One of the voltages in the BIOS was displaying at 1.2V but it was actually 1.5V+, killed the CPU.
>>101428454Very niche issue then
>>101429003Issue seems rooted in the CPU generation, not the version of software used. Also seems to happen more frequently in Datacenters with 24/7 uptime so it appears to be a degradation issue; the software isn't changing but the rate of crashes goes up in the past 6 months since the CPU was purchased.
>>101430120Bartlettlake-S 12 p-core is coming to lga1700 next yearThere'd be no excuse to getting an estrogen CPU then
>>101419538Intelbros.. this doesn't align with my deluded view, so it can't possibly be true. R-remember the 6 blown up 7800X3D because of a motherboard manufacturing issue.. AMDrones are so fucked
>>101419538your cpu on estrogen cores lmaoooooo
>>101430434goodlast real CPU was 10900k
>>101430533Ah yes, the modern Intel mobile experience. Full performance only when laptop is plugged in because only then it can reach the ridiculous turbo TDP. On battery they won't reach top frequencies, ever. Which is still better than whatever the fuck Qualmeme is doing with their 120ms ramp-up times on battery.
>>101419583I'm happy for everyone that doesn't have to deal with Visual Studio because of Intel avivs incompetence
>>101430533why do they lie like this? Intel is usually the safer choice because their verification team doesn't suck total nigger dicks, but that's only because the meltdown burning really helped them fix their shitty verification team out.
>tfw i9-12900kI'm so glad I didn't bother upgrading after accidentally my motherboard.
>>101431201They had to push their process to the absolute limit in order to fight AMD (and TSMC).>muh 6ghz stockturned out to be a grave mistake since the voltages required to sustain it also apply to the bing bus, and that's what is degrading in the highest end SKUs.
>>101431271>They had to push their process to the absolute limit in order to fight AMD (and TSMC).ya, but those are gaymer SKUs, not mobile slop bullshit or the lower bin'd non i9's.
>>101431278And it's the gaymer SKUs that are crashing, aren't they? Basically i7/i9 K and K(y)S editions?
>>101431304ya. Intel Aviv's verification team really dropped the ball on it and I'm not sure why. I really am not happy about it, but I don't think it's systemic to the whole product line.
>>101431304gamer skus are 1*100-1*600
>>101431329It's systemic to the whole product line in terms of the ring bus being vulnerable to voltage induced damage.However under stock condition only the highest end SKUs will ever request too high voltage in order to maintain high clocks, so it doesn't affect the rest of the line.On the other hand the reason might be something else, and everything is affected but the top is just now experiencing degradation with the rest to follow in 1-2 years.
>>101425357you should have bought some paper and crayons instead because you have to be retarded to invest in something you dont understandits "people" like you that made crypto kikes millionaires overnight
>>101431334Maybe for poors.
>>101431368yeah buy 12 ecores for gaming are you stupid?
>>101431381Who cares about cuck cores? They are parked most of the time. It's the ST frequency of 1 or 2 threads that matters most.
>>101431368if you think 700 and 900 SKUs are for gamers take the advice above and buy some paper and crayonseven for productivity a 7950X makes more sense due to its power efficiency
>>101431411so buy a 14600k, disable fag cores and set all core to 6.5 lol
>>101422927Yeah, class-action lawsuits are probably being prepared by lawyers as we speak.>>101422974>you'd need a new moboWouldn't a 7950X and a corresponding mobo not actually even cost less than top-of-the-line Intel CPUs?Of course they'd have to kiss their profit margins goodbye.
>>101431421>>101431425Just read Intel PR, they are targeting those at gaymers.Also why would I buy a housefire CPU when I have a 7800X3D?It's like you guys can only ever think in black and white. With complex opinions being anathema.
>>101431425Don't forget to undervolt, enable mitigations and disable hyper threading for a safe and effective Intel© computing experience.
>>101431459THERMAL VELOCITY BOOST
>>101422927>To be fair, what would they do, replace it with another faulty CPU?Offer either a full refund or a replacement of the CPU with a new revision that is fixed and doesn't degrade. Obviously the latter option would take a while, so in practice I'd expect most people to take the full refund and buy something else.
>>101431557in reality you don't get shit maybe a "$20 off your next Intel® purchase" coupon
>>101431567Yeah, probably. Glad I don't have one of those ticking time bombs on my hands.
>>101431567Or $4 in mail after 5 years because of a class action lawsuit.
>>101431356you gotta consider the low binned i7s would need higher voltage that can reach good binned i9 territories
>>101422656>refundsmight as well have Intel file for bankruptcy. they don't have money to refund 3 years worth of customers.
>>101432425>jews are suddenly bad at creative accounting if a condition that isn't "big line go up" arises
>>101432425tray cpus have a 1 year warranty
>>101419538Oh jeezas you had me worried then, I thought itwas something real men had to be aware of but looks like just a few kiddywinks are affected
https://www.tomshardware.com/pc-components/cpus/intel-next-gen-desktop-cpus-may-run-even-hotter-than-current-ones>Arrow Lake configured to run up to 110C before throttlingThey're still shoving power into the chips to try and compete, aren't they? Will they ever learn?
I used to believe that AMD were the poor man’s choice for processors when compared to Intel but it looks like they’re much better than Intel now, both in terms of performance and reliability. The only area where Intel can claim superiority is having better battery life
>>101434389the fact that youre provided good value for your bucks doesnt mean its a poorfags option.it would be like saying that makita is the poorfags choice when it comes to powertools (most construction pros in the west are gonna use makita bc its good value)
>>101434389>The only area where Intel can claim superiority is having better battery lifeBut they only get this by mercilessly murdering performance when on battery. So does it really matter that you get better battery life when the thing downclocks to 2ghz if not on wall power?
>>101423363No becahse your joke of a computer won't even try work fast on heavy loads
Librebooted core2duo gang rise up
Do you think this affects high end laptop chips too? Or just desktop chips?
>>101434653Nobody knows. But given that all the non overclockable chips and i5s aren't crashing, it's probably fine (for the lifetime of the product at least, if it crashes in 10 years nobody will care).
>>101434706The non overclockable chips aren't crashing because they aren't actually raptor lake, just alder lake. And we don't know that 13600K and 14600Ks aren't crashing, they could be but at a lower rate than i7s and i9s.
>>101419538Intel processors are now the greenest CPU on the market now that they're biodegradable.
>>101428057I mean 5800x is under 200 and goes up to 4,7. It's quite close
>>101435867it boosts to 4850 MHz if you have decent cooling, but you're obviously not getting that on all coresthe new 5800XT should boost to 4950 MHz (assuming they bumped the clock by 100 MHz) which is really close
>>101434706Intel haven't made many x900 non-K and x700 non-K CPUs since 9th gen so it's no surprise those aren't represented in the crash statistics
>>101434653>>101434706>>101434814>>101436307the absurd voltages required to sustain the 6 GHz boost also apply to the bing bus, and that's what is degrading in the highest end SKUs.
its pretty strange to have an outrage of this magnitude thats been going on for months since that nvidia "its not our drivers, its intel" and not have any real statements from intelI guess its difficult for them to suddenly start to investigate a problem that takes time to manifest in a chip in the field. Or they're just panicking and trying to figure out how to not crash the stock price with 2 gens of cpus being faulty and not performing up to what they marketed as.
>>101423384>laptops only boost to 157Ware you serious?
>>101438398you can read the specs on intel.com
>>101438309>Or they're just panicking and trying to figure out how to not crash the stock price with 2 gens of cpus being faulty and not performing up to what they marketed asit's this, i guarantee they found the root cause a long time ago but whatever it is (probably excessive voltage) would cause a massive performance loss if patched
>>101438309It doesn't take time at allSome chips are unstable under load from the factory like buildzoid's 14900k that crashes in r23 from day one, and that was a review sample from intel.
>>101426672No, it is ring-bus and E-cores being the weak link. They don't handle the voltages that boosting profile demands for near 6Ghz. Not a coincidence it occurring on SKUs with at stock boos profile that goes in excessive of 1.4V to reach near 6Ghz. It is on the same node as previous Alder Lake chips which barely cross 5Ghz barrier at best with more modest boosts which never had any of the issues.Intel literately resorted to hardcore factory overclock in response to Zen4 and thought they could get away with it.
>>101419563>Get real work doneLike...?You dumb fucking slave lmao.
>>101425357Should have bought AMD in 2016
>>101440509$1.25 per share......
>>101420312>intel CPUs degrade hardIntel:>hmmmm lets push T-MAX from 100°C to 105°C for the 15th gen to milk every bit of performance, sure it wont degrade even more
AMDbros just can't stop winning
>>101428117They always had parity for a while kiddo on the desktop front. On the workstation/server front AMD has a power gap in I/O connectivity.
>>101419563You fucking wagecucks are insufferable
>>101440555>They always had parity for a while kiddo on the desktop front.Intel's CPU uplink is 8 Gen4 lanes wide, AMD's is 4. You literally can not max out a Gen4 NVMe SSD on AMD chipsets, because the bandwidth is shared with everything else. Only the next generation chipset will finally reach bandwidth parity, by upgrading to 4 Gen5 lanes.
>>101419538>Laughs in AMDI wondered when shit like this would start happening
>>101440548>do nothing>winBASED LISA "TURNING INTEL INTO INCEL" SU
>>101440460Yeah 12900K's maxed out around 1.35v for 5.2ghzIntel went full retard with 1.46v at 5.8ghz , 1.5v at 6ghz and 1.55v at 6.2ghz on the latest gens of cpu
>>101441226blame /twg/, the worst general on this board.
>>101422656then class action sue them.