funny thing about that comment is he was refering to GN, Steve did say how good of a job Steve did benchmarking different profiles, so Steve did credit Steve, back to you Steve
Don't forget that Intel also denied warranty to anyone admitting to using XMP by claiming it as overclocking (even though XMP has been sold by Intel as the default for all Intel certified RAM kits for at least the past decade and a half).
@@NJ-wb1cz Intel said the CPUs would randomly ask for 1.6v, regardless of voltage settings. That would _absolutely_ cause an instability and likely damage over time. Probably part of their stupid to the moon boosting algorithm that's only effective for seconds. The i9s should still be in the same realm of performance without asking for 1.6v post-patch. The real question is how many CPUs have already been damaged and users just don't notice it yet?
@@zodwraith5745 I wouldn't trust Intel's words too much at this point, they are doing PR and damage control. My point was, _regardless the causes_ , one of their primary goals has to be to make as many of those damaged CPUs as possible work in a somewhat stable way so that Intel isn't overwhelmed with warranty claims and isn't bogged down trying to replace every damaged chip with new ones instead of producing their 15 lineup. And they can't do that without degrading performance
@@NJ-wb1cz Well, we'll know when we get the update won't we? If they really do only fix it so it doesn't ask for 1.6v as they claim then there will be little to no difference in performance. My CPU actually boosted _better_ after I undervolted it. If we see a notable drop in clock speed I'm sure channels like HUB and GN will hold them fully accountable so it's not like they're going to get away with any shenanigans, so the cause is _very_ relevant. Any issues that arise from this will have zero effect on Arrow Lake as that's an entirely different architecture on entirely different nodes. Again, you're making baseless assumptions making it look like you have an agenda. I'm responsible for building more than a few of these so I only care about the customer and Intel making this right. (Luckily neither mine or any other i9 or i7 I've built has had a hiccup so I already question how many are _really_ effected.) Obviously we can't blindly trust Intel's words, but automatically condemning them and assuming the worst as you're doing is just as foolish and reeks of fanboism. I don't trust AMD, Intel, or Nvidia, but I'm not dumb enough to automatically assume they're out to screw every customer like you are. AMD took _18 months_ to finally release 300 series drivers for Zen3 after they promised them and I never saw their fanboys complain when they _should_ have. You need to at least be fair and give Intel a chance to fix it. If they gimp the hell out of chips and refuse RMAs _then_ you can light your pitchfork and sharpen your torch.
Old cpus don't cos they're pretty much invulnerable to electromigration cos all the features in them are relatively large and they run at lower clock speeds. Current gen ones have such small features that it doesn't take much electromigration at all before their performance is affected.
It is a scary thought indeed when you think that average people don't follow tech news, how many Intel owner out there having instability issue and chalk it off as xmp problem, blaming motherboard, or even gpu driver..
Those people should be ashamed cause Google searching should be muscle memory. It's not like the 1900s where your only source for global level of information is newspapers or radio. Also if you own an enthusiast proc and don't take care or time to research your hardware says a lot about that person rather than the manufacturer
@@diamondlion47Intel is also better than Intel, Core i9-13980HX laptop is way more power efficient than a desktop i7-13700K on Cinebench R23 and can be faster (it beats a 12900K anyways). The perfomance per watt is ridiculous, i watched in real time how it performs in Baldur's Gate 3's city and absurd framerates without hearing fan noise from the laptop (Asus Strix 2023 with a mobile 4070).
@@Aggrofool And everything he sees is just blue like him, inside and outside. Blue his house, with the blue little window, and a blue Corvette, and everything is blue for him, and himself, and everybody around 'cause he ain't got nobody to listen to
I absolutely loved the garlic talk. Had me cracking up. My anecdotal evidence: my father (and I) always ate copious amounts of garlic growing up and essentially didn't need bug spray to keep the ravenous mosquitoes and black flies of Maine away while everyone else got eaten alive. Thinking back, I think the sickness resistance was also there for me. Marrying my wife cut down my garlic intake and I've definitely gotten sick more than I used to, but I've also gotten old (42 now) and have three kids including a 2 year old in daycare, so who knows. Time to up my garlic and see what happens 🧄🧄🧄
I also eat a lot of garlic, but mosquitoes annihilate me! Garlic has vitamin C which helps your immune system function. As well as magnesium and niacin, with have a massive variety of benefits for your whole body.
Before my 40s, I, too, rarely got sick and my symptoms typically lasted only a few days. In January, though, I had cold symptoms that just would not go away, even after several weeks. I finally bought a pack of Airborne immune boosting tablets, which are basically just high doses of vitamin C, and my symptoms went away within a day. I concluded that I must've been vitamin C deficient, so I've been eating a couple of vitamin C gummies per day since and haven't been sick again. So, instead of garlic, you might want to just try taking a vitamin C supplement every day and see if that's enough to get your immune system back to how you remember it.
I work for a huge company in the UK. We have a large amount of IT equipment (laptops, desktops, phones, tablets, etc). The companies policy is such that unless the item automatically bios updates then the device never has a bios update. We have a large IT department located in another country who are unaware of the Intel problem. At what stage is the mainstream public going to be informed? I only know because I watch This channel, Steves channel, etc not because of news items on mainstream news platforms
26:25 I think it’s the Australian accent, other people hear us putting an R on the end of random words when we don’t hear it. There are videos about the Aussie phantom R with examples 😂
Seeing the data coming from Wendell, Level1 Tech, combined with other reports, it is established that failure is always associated with high demand communications with connected devices: RAM, NVMe, GPU - and Wendell even showed that using four sticks of RAM was far worse than using two. So it seems obvious that either the Ring Bus is being overdriven, or the IO is overloaded. From 12th Gen. to 13/14, the only real design change was to increase Core Cache size. So it appears that they're asking too much of one or both of those data channels. Assuming they aren't lying about manufacturing defects.
Do you think with Gen 15, Intel addressed this possible issue? I really am ready to build a new pc after 10 years and Arrow Lake looks really appealing.
@@sirmonkey1985 well it should be a new architecture since it's using a smaller process Node and Gen 15 is using chiplets. I hope someone at Intel has the same though process as you and overhauled the Ring Bus.
Funny thing is they don't have to. They could literally put out CPUs that would blow up in a week and people will pretend like it never happened. Remember what happened with ryzen 7000? CPUs and motherboards burning up? Yeah everyone forgot about it. Come back in 5 years, people will still remember that a handful of intel CPUs are crashing using the stock boost algorithms. Whoever makes the cheapest cpus will always have the hand up, at this moment its AMD. People will fabricate any excuse and put that as the reason why they are not buying the 14900k when in reality its just that they are literally broke af.
@@chakcrak933 except that was motherboard vendors going outside of spec and AMD quickly offered a solution, which isn't like the Intel situation at all
@@chakcrak933 Everyone has their own reasons not to buy 14900K. If any of my friends were going to buy that cpu, I would whip that friends ass. Dead platform with problems, and new platform coming later this year. Use your common sense.
@@ChristmasCrustacean1 Intel *blamed* motherboard vendors, and there may be some truth to that, but as more details come out it appears that the motherboard "overpower" issues were an exacerbating factor at best; the chips still allowed themselves to operate in ways that were detrimental to themselves.
I had a 13900kf for about a year now. With stock settings turning off all the limits. It reaches 96c on a custom water loop. And multiple cores throttle. The newest motherboard bios for asus that ensures intel standards did not fix this. The only fix for this was undervolting the cpu - 0.0700 volts. Now no more throttling and a higher cinebench score. max temp is now around 86c max. But i still get a issue at startup where the cpu spikes to 1.5v. These issues are ridiculous and should be fixed asap.
The problem for Intel is that even if they replace 13th and 14th gen cpus with 15th gen cpus without the faults, the owners are still out a motherboard.
People need to file FTC complaints and state AGs if they are affected by this. Government has been functional under biden, so the ftc has been going after businesses cheating customers. They need complaints to act. There is no reason they cannot use their stock of the same chips, update them, and send them as replacements. (when the microcode is ready)
Intel should refund the mobos together with CPUs, seeing as how that mobo will be practically useless without it, and there's likely not going to be anything to replace the chip with. Of course, they won't, and it also will do anything to weasel out of replacing the chips as well.
Pretty sure that Intel knows why the CPU's were degrading the issue is that they don't wanna deal with the already degraded CPU's cause apparently they have been denying RMA on a lot of them before Wendell and Steve released their videos so if you have an already un-stable CPU what ever fix Intel is releasing will likely not be fixing your issues the fix that is coming is to stop the degradation and not a fix for those CPU's that have already degraded.
The lack of real journalism in the tech industry is the only reason these companies do not feel the need to respond. They have to be forced to make changes, WE ARE THE CHANGE.
@@TheHardwareUnboxedPodcast I agree, I just think the bar is really low because of conflict of interests...I bought a gamers nexus hoody the same hour the Asus video dropped to support them, would do the same for yall too. I have in the past supported both of you via patreon.
@@marktackman2886 Gamer's Nexus, Hardware Unboxed, and Level1 are the the best of breed for accurate information without BS speculation and nonsense. I won't even bother with LTT anymore because I think they have, or are prone to having, conflicts of interest.
(Removed details about my inept first attempt at undervolting, lol.) Games have quit hitching, so that's ...nice. Cool stuff, Intel. Mad respect to AMD for stepping in and preventing an Intel-scale CPU debacle. Great PR move for AMD, too. AMD does what Intel isn't Intelligent enough to do. Doesn't quite have the ring of Sega's slogan.
That didn't helped my unstable 14900K, what it turned it back to stability were, lowering down the voltage to 0.85v and capping freq to 5.5ghz. I still have PL1 at 253W and PL2 to 288W with ICC 400A, and now it is pretty rock solid even if it ended as a nerfed down 13900T
@@justdointhisforthegames certainly i just disabled all thermal and voltage optimizations on my msi x790 pro and it dropped down automagically from 1.2 to 0.85v, not modified any offset or did manual undervolting
@@Pacho18 neat trick. i might try something like that. brb while i brick my pc, haha; glad it worked for you but i don't trust my motherboard that much
I am 13900k limited to 262 watts, 1.28v undervolt, load line level 6, cores pinned at 5.5/4.3 on p and e cores. And not 5.8 ghz boosts. Been stable since beginning, not hitching. First chip degraded even with low VID voltages, but first cpu was limited at 300 watts. With Multicore Enhancement ON it will draw 350 watts or so. With it off it hovers around 255 watts or so in benchmarks. 2nd cpu limited to 262 watts is fine. You doing 125 watts is beyond hilarious. You might as well stick 10900k at those downclocks on your cpu at that point and 10900k probably will win.
A minor correction to your statements around the 59min mark, where you said that AMD was nearly bankrupt because their processers weren't very good. You're missing an important detail, in that the years prior to that saw Intel literally paying vendors NOT to sell AMD products. There was definitely some mismanagement on the part of AMD leadership (like acquiring both a fab and ATI when their budget definitely didn't allow for that), but the driving force was Intel blocking them from selling their products to OEM manufacturers. This led to the decline in their CPU performance, which led to the David/Goliath situation we saw between AMD/Intel.
UntiL AMD demonstrates the same level of systematic sleaziness seen with Intel, I'm not going to place them in the same group. The current lawsuit-friendly shady behavior by Intel regarding this serious issue is in a league of its own.
I like the freedom to destroy things I buy through my own stupidity. Manufacturing screwups or out of control parameters I’m unaware of is an issue. I’ve got a 7950X3D I’m all for a Bclk overclock to get better than 5.25 on cache cores. As long as you know all your voltages are safe running delidded w/custom loop like I am, crank it. I
Blaming Board Partners ignores the fact that ALL of Wendell's data comes from Server boards that have ZERO overclocking and are configured to the minimum specs for these CPUs.
someone said they were running 6ghz 1.5V around 65w. because the server was running like 2 cores so the boost algorythm kicks in. and btw it boosts that high without any extra overclocking
Allicin is not super effective, but it's highly reactive, very nonspecific (meaning it works against a wide array of bacteria), and short-lived, which means lower chance that bacteria get resistance to it.
A strategy with dedicated game servers is to use a LOT of relatively low cost machines(relative being a key word). Not only does this limit the blast radius when servers go down, this strategy also keeps latency low, since you can distribute these servers all over the place. It would also be cost effective on a 'per thread' basis, assuming the chips are stable that is. But intel's issue has been a tough one in that space. i7's and i9's, 13th and 14th generation carry the plague and are dropping like flies. A lot of datacenters are buying up 12th gen cpus, but those aren't stocked anymore, so are being forced to pivot to Ryzen cpus.
I genuinely can't believe people were still buying Intel. They were MORE EXPENSIVE while providing less performance (on top of higher power usage). I guess brand loyality is one hell of a drug
You are doing the same what you accuse other people of doing. If you want to jump on the bandwagon that Intel is an evil corporation and poor old AMD just wants the best for their customers...do me a favor.
Brand loyalty might have a lot to do with brand and platform familiarity. Not everybody has had the opportunity to experience both brand. Motherboard prices and features play a role as well. When I bought my z690 and 12600k with good cheap ddr4000... 5800x3d wasn't out yet. At the time it was cheaper than the 5800X and if I wanted more features then the motherboard price was just ridiculous. My overclocked 12600k pulls max 120w in gaming but arround 200w in cinebench R23 with 19600 multicore score. 5800X3D and 7800X3D with cheap ram annihilates my setup in gaming but on the multicore rendering ? It's safe to say AMD is better but that doesn't mean Intel cpu is worse on all front. I'm much more concerned with the 1% and 0.1% lows and yeah the X3D are the cpus for that.
I think it's kind of like how people make a sports team part of their identity, even though they have no actual connection to the team at all. That team must be good or they feel less good about themselves.
only way i can see comparing zen 5 to 5800x3d is that it's most probably most probable 2 generation upgrade cadence, i doubt 7800x3d people will be looking at 9700X or 9800x3d, at least not majority
There's a leaked photo showing a Typo on the Heat Spreader of a 9700x. Basically it seems like a bunch of Ryzen 5 and Ryzen 7 CPUs were mistakenly labeled as Ryzen 9. Apparently that's the QC issue causing the delay.
Intel CPU issues on the server side should have all the companies relying on them looking into sharing telemetry to discover these potentially business harming problems sooner in the future. Wendell has shown the data was there, just no one was talking to each other.
Yeah, just waiting for the other shoe to drop here. Intel's problems are certainly much bigger, how much remains to be seen. I'll be staying away from their chips for now.
My 13700 laptop started having stability issues after about 4 months. I undervolted and put kryosheets on cpu and gpu. It fixed it for now but my avg fps definitely took a hit.
@saricubra2867 I wouldn't say the only thing. It was quite noticeable in the game I was playing at the time (hogwarts legacy) where I went from 70's avg fps down to about 60. My 1% lows stayed about the same.
a delay is soon forgotten, suck is forever, AMD is doing the right thing by making sure that we get good chips that operate as intended, unlike Intel and what they are going through right now.
So I have an I7 14700KF. Until the Microcode update am I safe with the Intel Default enabled in BIOS? Or should I also not gaming until then to not damage the CPU? I don't have stability issues. Thanks!
Intel Default is not the best. It sets the AC Loadline to 110 or 1.10 (depending on how the motherboard handles it), which cranks the voltage incredibly high to the point of potentially damaging the CPU. The best thing you can do is disable the single-core boost as it is fairly meaningless on a desktop system, and only serves to win benchmarks and degrade the CPU faster. After that, just undervolt the CPU and that will ensure you get the most out of the CPU. TLDR: Disable single-core boost and undervolt via lowering the AC_LL.
My opinion is that Intel are attempting to covid-style "flatten the curve" by reducing the rate of damage via soft limiting while they can build enough stock to stuff their, and OEM's warranty/recall supply chain. Fact is, they simply cannot do an instant recall. There aren't enough parts or technicians in the world to declare a no questions swap out. With desktops and servers it's relatively easy. Build and deliver the parts, pay ~30min swap labour. With laptops it's much more of a nightmare. OEM's have to build replacement motherboards and fill their supply chain - and Intel has to reimburse them. Now, it might be possible to only build 20% more boards and de/resolder swapped boards, to achieve 100% replacement, but it's still an expensive proposition. The best they can do is slow down the return rate while they implement an effective recall process with all their partners - something they haven't had to do for a long time. The institutional knowledge of the process to achieve this is long gone and/or obsolete. The other problem is that errors are less desirable than reduced performance. It may have reached our attention because of gaming, but gaming crashes are merely annoying in the grand scheme of things. These processors are deployed in mission critical roles performing all sorts of tasks where errors have real economic and potentially human costs. Financial systems cannot abide a processor that can't math accurately. Safety and medical systems can't either. Imagine an xray machine comparing the $currentradiation to $maxradlimit and giving the wrong result. It's remote, but someone could plausibly be irradiated. Even a billion to 1 error will happen thousands of times a day across the deployed fleet. "flattening the curve" is their safest option right now.
Dude! Garlic is very well known for improving your immune response like this. I personally hate it and avoid it, but even I one time couldn't get past a really long cold and started eating garlic in addition to taking my meds that just weren't enough. That solved it.
Also it was obvious in the test samples they sent to reviewers for review, that they had a issue/s, but they haven't revealed what it is yet as of 2 days ago now. Speculation wont help unfortunately here.
i may sound like an AMD fanboy, but i feel like there is just more good will in AMD, all those open source projects, all the things you guys mentioned, 7800x3d melting issue, they did cover for Asus error, while year later Intel throws all board partners under the bus AMD does have a wierd marketing team, which quite often makes things awkward, but they are company, profits is all that matters, but from my point of view, in the end one thing matters for a consumer, how they solve the problem for example, Nvidia blocking FG behind 4000 series for artificial reasons, AMD could do the same for AM4, just say that zen 3 needs some hardware tuning in chipsets, so only 500 chipsets can work properly and push that, some will say they crumbled under the pressure, but in the end they just did the good thing, same with FSR, it being open source gave a lot for GTX and RTX users who wants Frame Gen or plain old upscaler, Intel also got a lot of praise for XeSS being open Digital Foundry made a clip about AMD research about hardware RT and Meshlet technology and both of them are not locked by any hardware like Tensor Cores, so i think that kind of openess gives the feeling of being trustworthy, but in the end, yes they are a for profit company and have share holders to satisfy
@@roboninja1 not in GPU for sure, but in enthusiast desktop CPU and servers, i believe they are, better or same performance, more power efficient, no need for industrial grade cooling solution, cheaper, 7950X was what $100 cheaper than 13900K/14900K? Intel strong side is mixed workload gaming and professional on i7 parts, other than that it's mostly within 5% difference, so kind of nothing soup, you'll be happy from either of those as long as they are stable thing for system integrators is like someone said "you don't get fired for going Intel" they were golden standard for so long, that it was just safe bet, but now with all those problems it will stir the pot a little and locking features in doesn't take much, it's even easier since there's no need to make it accessible as much
The Intels in my past: 8086, 386, Pentium, Pentium MMX, Pentium III, 6700K. The AMDs in my past: 286, 386(40), 586(133), K6-III, Athlon MP, Opteron, Phenom II, Bulldozer (Vishera), and now AM4. The fact that Intel started requiring new motherboards each year ended it for me. I have an awesome motherboard for my 6700K that should have been sufficient until at least 9th gen. My rule of thumb is if the PCIe generation doesn't change then you shouldn't require an upgrade.
@@1Grainer1Raptor Lake mobile is lightyears better than desktop. The Core i9-13980HX is extremely impressive, it's faster than a desktop 13700K and more power efficient. The perfomance per watt is also better than a 13900K/14900K which is funny because it's the exact same sillicon. A friend has a Ryzen 9 5900X desktop tower and he uses that i9-13980HX Asus Strix laptop for work during travels. Crazy perfomance in Baldur's Gate 3's city (which is extremely CPU intensive) and i didn't hear any fan noise.
@@1Grainer1 Huh? There has _never_ been a point in time that the 7950x was cheaper than the 13900k. 13900k launched at $590, 7950x launched at $700. 12900 launched at $590, 5950x launched at an eyewatering $800. 10900k launched at $490, 3950x launched at $750. Sometimes the price gap is big enough to straight up buy a new motherboard. Even right now the 13900k is $440 while the 7950x is $525. You can complain about efficiency and thermals and that's more than fair, but Intel is flat out _NOT_ more expensive than AMD. AMD has been going for your wallet for years now. Anyone that claims otherwise is a straight up fanboy spewing copium, or simply doesn't know what these parts _really_ sell for. The fact you magically think the Zen3 on non-500 boards wasn't an issue is pretty telling. Especially when 400 owners had to wait 4-6 _months_ for BIOS after launch, and 300 owners had to wait a laughable *_18 months?!_* That's not doing the "good thing". That's a straight up backhanded tantrum that they were pressured into offering compatible BIOS. That's the kind of "compatibility" Apple gives when the EU sues them for anti-consumer practices. _EVERY_ time AMD has been in the lead or even matching the competition they've charged sky high prices. They did it with Athlon, they did it with FX and their 8 cores. (The ones they got sued for?) They've done it with Ryzen. And how's that Threadripper pricing today? AMD is NOT your friend and NOT the "good guy" as much as you think they are. They do it with CPUs, motherboard chipsets, as well as GPUs. (Remember Fury pricing? Vega pricing?) @roboninja1 absolutely has a point that they only get honest when they have competition. _Exactly_ like Intel and Nvidia. That's not to say Intel and Nvidia aren't just as shady, but to pretend AMD is somehow any more noble is just plain naive. I daily drive AMD and Intel systems and I hate ALL these companies. But it's a straight fact that my one AMD system has had FAR more issues than the 3 Intel systems in the house combined, and it wasn't the cheapest one either. And yes, one of those is 14th gen. I would definitely say your fanboy glasses might be clouding your judgement. F Intel. F AMD. F Nvidia. None of these big evil corporations deserve your fanboism. It's brainless sh*t like that that let's these companies think they can get away with the BS they pull on a daily basis.
If the 13 and 14 CPUs are not fixable, and the replacement option would bankrupt the company, another option is to issue a credit based value to be applied to the future (next) gen CPU. This way not having to layout a cash loss with direct replacement. Regardless it is serious hit on the company not just now but going forward as well.
Hi guys. I’m currently shopping for a work laptop upgrade. Would you suggest i stick with the i7 14th gen instead of the i9? I’m purchasing in the coming two weeks as I’m still researching.
I'm still on the fence : am i in deep sh*t with my core i5 13600k? I own that CPU since january 2023. It is installed on a Gigabyte Z790 Aorus Elite AX DDR4 mobo. My casing has plenty of airflow (EATX Fractal Torrent) and my cooler is the great Noctua NH-D15S Chromax Black. The fan curves are fixed in a way that my games rise the CPU heat at 69-74 , max. It's not overclocked and I dont know how to undervolt it If it's required to give it a better lifespan. I watched some videos out there but none of them were clear enough for the casual users like me. I need a clear newbie guide step by step. I downloaded the bios update, extracted it at the USB key root. When i boot my PC with it, the boot logo show on screen but nothing else ever happen.
I do QA at a software company, at first our clients had full access to the whole software, then we had a new CEO and he decided we need to put limitations so that the clients pay more to have those limitations removed, which is a practice I really hate, so anyways, he said it so we had to implement it, then while doing QA I noticed that there's a way to bypass the limitations, I validated the change and told nobody about that, although I was almost sure none of our clients would know how to bypass the limitations. Sadly, 3 months later another dev discovered the bypass and patched it. Right now they want to prohibit account sharing to force the clients to pay more, I am against it but there is nothing I can do about it. Also it's pretty hypocritical because we do share accounts on some services we use. So yeah anti consumer decisions do not reflect the employee's mindset, but rather the top level management's
My more-than-a-decade-old Intel Core i3-3220 is still kicking around just fine in my old PC, and here I am, worried about my 4 months old 13600k degrading over the next couple of years, I was planning to use to for at least as long as my previous PC but it seems like it won't even last half as long as that one. If Intel can't fix this problem, then Intel should be sued to the ground for this unless they provide us with refunds and compensation for buying into their platform/motherboard that only works with these defective CPUs (technically it'll work fine with 12th gen but that's a performance regression so it shouldn't count).
The notebook cpu issue it's underrated. A desktop cpu fails, you can at least replace it, at most together with a mobo. It also has a 3 year warranty. A notebook CPU dies, many notebooks have warranties of 1 year, 2 years tops, and mobo+cpu replacement, even if available, is usually as expensive as buying a new notebook
The garlic story actually make sense somewhat. In Greek culture garlic is believed to grant eternal youth or something like that, it definitely has some positive health related reputation in tradition. Thats why the Greek put copius amounts of it in food. I'm not even kidding
55:19 they definitely are becoming more anti-consumer. I wouldn't be surprised if B840 is just a first step in increasing the average sale price by slowly eroding the B-series until it becomes a hollow husk with none of the features that made B-series so good.
I bought a 13700K in late 2022, so it seems likely mine has the oxidation issue. If that's the case, there is no fix other than replacement. But so far I've had no issues. Important question: what's Intel going to do for customers who have this ticking time bomb that hasn't exploded -- yet?
Does the 9th gen i9 9900k have the same issues of degradation? I can not run mine on turbo with hyperthreading. Can't clock memory to 2666 or system freeze will happen in windows.
You get a lot of vitamins and minerals from garlic. If you wife can't eat garlic, I'd recommend golden kiwi fruit, as they're high in Vitamin C, and are much nicer than regular kiwis. Or any citrus fruit really. Just avoid replacing fruit with fruit drinks, as they often lack the vitamin C, as it breaks down in the production process, and is often flavored with massive amounts of refined sugar, which is obviously bad for you.
I can’t begin to imagine the average user opening their pc and replacing the CPU after removing the attached cooler. So not only is this a part replacement issue but a huge labor and shipping cost.
Buildzoid had an interesting video also referencing Igor’s lab test on cpu voltage and it seems due to binning that certain CPUs are basically running stock over voltage and others have much lower voltage. Some were stock almost 1.5V and others stock 1.3 and less. This on both i5,i7 and i9. So your mileage may definitely vary. But if your cpu is stock running close to 1.5V I assume that is definitely going to impact longevity.
I am OK with the delay. I have a 7700X in my SFF build and I am not looking to upgrade at this time. I think AMD is doing the right thing by pushing out the delay to make the launch successful.
Steve interim project suggestion; Sell me on DeWalt. I've been a Makita man for a while. Had a bad experience with a DeWalt drill (corded) so I've been a little nervous about them. Give us more some Hardware Toolboxed.
I think the reason for AMD delay will be visible on release notes for new BIOS releases for motherboards that already released their Zen5 BIOS-es prior to launch. That is if something had to be updated via microcode update.
I really hope it doesnt affect the 13900HX (mobile cpu) where it is a lot more difficult to replace. Lenovo recently replaced my Legion 7i Pro motherboard due to an issue where the symptom was the CPU reaching 90-100 degC at idle. This issue popped up after less than 12 months of ownership. While there are other potential explanations for this, it was really odd due to how it came about. Fingers, toes and doodle crossed that this insidious problem isn't the cause as it is a 4 grand laptop.😢
They give raw garlic too to cows to not pump them full of antibiotics is a good remedy an not snake oil Btw you find over the counter raw garlic in flexible shells so yo can skip the bad breath and all
meanwhile on intels side "Sir we can not compete with Ryzen" and the boss so "Shut up and sell them for double the price", Worker "But sir the Products haveing issues" and the boss "So shut up and increase the price again so we get the most amount of this crap".
AMD was (still is?) in similar place with the Zen3 WHEA error 19 debacle, and to some extent with the X570 USB disconnect problems. I had to return my 5950x which was not stable in light threaded workloads out of the box, and I was far from alone in this boat.
To add some further mysticism to the wizardry, I would add a past gf who was Chinese very much believed in traditional medicine and had me drink ginger garlic tea when I would get sick and it helped me recover every time. Garlic has anti-microbial properties. You can look into it more to get the proper explanation so Steve's story makes sense to me. We ate a lot of garlic in general for that reason. Chew on some raw garlic for a sore throat if you can handle it
seeing how AMD handled the x3D chip failures i'd like to think they'd handle it better but agree it's hard to say exactly what they'd do given how much of the customer base it effects(OEM/server/DIY) and wouldn't surprise me at all if they ended up doing exactly what Intel's doing right now. but i also think the way AMD handles the bios limits and restricting board partners on what they can and can't change is done a little better.
For people using 13th/14th gen cpu's what would you recommend they do/check to see if their cpu is degraded or having issues but not dead yet... I do not expect Intel to really help end users out
"optimistic" binning assumption is reasonable since their chip making division is running a net loss, so they're stimulated to sell more high-margin SKUs
My TUF Z690 board has a new BIOS released this week from ASUS, stating to have the 0x125 microcode fix. I've got a 12700KF so I don't know if I "need" this update.
The 12700KF is so power efficient that at stock it uses low wattages and voltages. It's literally better than the entirety of desktop 13th gen and 14th gen K chips.
I have a 13900k with 2 sticks ram (xmp on) at 6400. CPU cooler is 360 Deepcool AIO. My system was made around late December 2022 haven't had any crashes OR bluescreens. Am I good how do I check if I have degradation?
Buildzoid made an interesting observation about e-core frequency boosting causes the chips voltage to go over 1.6v. I'd try disabling the e-core boost and see what the effect is, and what type of performance differential it would create.
Is it not illegal for Intel to refuse you a RMA if you refuse to “down clock” your cpu? You infact paid for advertised speed and a down clock is below the speed you paid for?
It’s a huge mistake if Intel has already been selling a lot of these chips at near-zero profit margin simply to block AMD from marketshare this whole time, and this problem really shows why. If you already didn’t make money on the original sales and now need to replace parts for free, well…now you’re shedding money for no reason. And then speaking to what’s being articulated roundabout midway through, if you RMA the hardware without having a software resolution to adequately protect the replacement hardware you’re just heading for another round of RMAs down the road. Intel’s hope definitely has to be hoping this can largely blow over after ArrowLake releases to minimize RMAs they have to honor, and they have to accept doing ArrowLake properly beforehand this time even if they have to cede the performance crown another generation. There are so many levels on which this entire problem has been mismanaged and it’s really putting Intel in the hotseat. Intel needs to acknowledge internally they aren’t in a position to continue pretending they’re in the lead anymore and simply compete where they actually are.
24:32 Yes, many people say bye-ose. Most, by my reckoning, but I haven't done a survey. Saying bye-oss sounds strange to me and many others, so that may be what's going on here. Probably just a regional pronunciation thing.
According to Moore's Law is Dead latest video, the oxidization issue occurred in their Chandler AZ Fab between March and June 2023. We still do not know if those CPUs made it out to the channel or batch numbers that the end user could check to see if their 13th/14th CPU are from these batches.
Kinda interesting to hear handball referred to as obscure - but I suppose for Americans and Australians it is. Oddly enough, I think it could be very well received in both the Americas and Australia, especially with some minor adjustments to rules and aesthetics. I imagine it's more of a coincidence than anything else that it isn't a thing in those areas.
Honestly my theory is that the oxidation issue is the main issue but since that is completely un fixable for existing cpu owners they throw the voltage excuse out there as a possible xause as a least worst explanation that you can get a microcode fix for. Then when cpus fail anyway, people just assume it was too late that the microcode fix got applied and they were just unlucky. It isn't as if every cpu owner has a materials analysis failure lab to call bullcrap. The fact that they said the oxidation issue was only for a period of time and got fixed back in 2023 but they have not posted a website with all the affected cpu serial numbers for recall for an apparently known date range of affected cpus. That is highly suspicious that the issue was actually for much longer than they are letting on.
Someone needs to get this issue on the big, mainstream news channels. Imagine listening to those pundits talk about "Intel CPU Instability" and "Microcode patches", and basically everything in the Verge article published just before noon yesterday. That would light some fires.
funny thing about that comment is he was refering to GN, Steve did say how good of a job Steve did benchmarking different profiles, so Steve did credit Steve, back to you Steve
Steveception 😂
Thanks Steve
The more you Steve. The more you Steve
All glory to the Steveperium!
its just Steves all the way down
the cherry on the cake is intel saying don't run your ram so fast after telling fast ram was a feature for 2 years
👌🏻
But their fanboys will still say AMD don't support more than 6400Mhz lmao
Don't forget that Intel also denied warranty to anyone admitting to using XMP by claiming it as overclocking (even though XMP has been sold by Intel as the default for all Intel certified RAM kits for at least the past decade and a half).
I imagine that "replace all the affected CPUs" is just above "dynamite the building and flee to the Caribbean" on Intel's list of options.
Well those chips will be replaced regardless, by AMD 😆
Which is why the patch will likely slow down CPUs to try to recover as many of those burnt out ones as possible
@@NJ-wb1cz Intel said the CPUs would randomly ask for 1.6v, regardless of voltage settings. That would _absolutely_ cause an instability and likely damage over time. Probably part of their stupid to the moon boosting algorithm that's only effective for seconds. The i9s should still be in the same realm of performance without asking for 1.6v post-patch. The real question is how many CPUs have already been damaged and users just don't notice it yet?
@@zodwraith5745 I wouldn't trust Intel's words too much at this point, they are doing PR and damage control.
My point was, _regardless the causes_ , one of their primary goals has to be to make as many of those damaged CPUs as possible work in a somewhat stable way so that Intel isn't overwhelmed with warranty claims and isn't bogged down trying to replace every damaged chip with new ones instead of producing their 15 lineup. And they can't do that without degrading performance
@@NJ-wb1cz Well, we'll know when we get the update won't we? If they really do only fix it so it doesn't ask for 1.6v as they claim then there will be little to no difference in performance. My CPU actually boosted _better_ after I undervolted it. If we see a notable drop in clock speed I'm sure channels like HUB and GN will hold them fully accountable so it's not like they're going to get away with any shenanigans, so the cause is _very_ relevant.
Any issues that arise from this will have zero effect on Arrow Lake as that's an entirely different architecture on entirely different nodes. Again, you're making baseless assumptions making it look like you have an agenda. I'm responsible for building more than a few of these so I only care about the customer and Intel making this right. (Luckily neither mine or any other i9 or i7 I've built has had a hiccup so I already question how many are _really_ effected.)
Obviously we can't blindly trust Intel's words, but automatically condemning them and assuming the worst as you're doing is just as foolish and reeks of fanboism. I don't trust AMD, Intel, or Nvidia, but I'm not dumb enough to automatically assume they're out to screw every customer like you are.
AMD took _18 months_ to finally release 300 series drivers for Zen3 after they promised them and I never saw their fanboys complain when they _should_ have. You need to at least be fair and give Intel a chance to fix it. If they gimp the hell out of chips and refuse RMAs _then_ you can light your pitchfork and sharpen your torch.
You heard it here first. Steve says to use liquid garlic as thermal paste and the intel issues will be solved. #science #gandolf
Fking gold man lolz
*Liquid Garlic Thermal Paste a day keeps Vampiric Degradation and Oxidization at bay* 😂
CPUs don't normally die within 4-5 years. My AMD Duron 750Mhz from 25 years ago says hi 😉
Old cpus don't cos they're pretty much invulnerable to electromigration cos all the features in them are relatively large and they run at lower clock speeds. Current gen ones have such small features that it doesn't take much electromigration at all before their performance is affected.
my amiga 30 year old
@@spankeyfish The flip side is that older components pushed a LOT more voltage through to keep the circuit alive.
My Pentium replies to your machine and says h
now updated it says hi
It is a scary thought indeed when you think that average people don't follow tech news, how many Intel owner out there having instability issue and chalk it off as xmp problem, blaming motherboard, or even gpu driver..
Those people should be ashamed cause Google searching should be muscle memory. It's not like the 1900s where your only source for global level of information is newspapers or radio. Also if you own an enthusiast proc and don't take care or time to research your hardware says a lot about that person rather than the manufacturer
What I've taken from this podcast:
Put garlic on 14th gen 🧙♂️
I took away: buy AMD 😆
@@diamondlion47Intel is also better than Intel, Core i9-13980HX laptop is way more power efficient than a desktop i7-13700K on Cinebench R23 and can be faster (it beats a 12900K anyways).
The perfomance per watt is ridiculous, i watched in real time how it performs in Baldur's Gate 3's city and absurd framerates without hearing fan noise from the laptop (Asus Strix 2023 with a mobile 4070).
@@saricubra2867too bad that CPU is degrading since it's a rebadged desktop part
@@asdf_asdf948 It doesn't reach desktop i9 voltages.
To be fair it’s wizard approved garlic. I trust wizard doctors as all should.
AMD blue balled Steve 😂
Bastards! Just because I said their benchmarks were BAD!
Intel is Team Blue
Yeah they did 😂
@@Aggrofool And everything he sees is just blue like him, inside and outside.
Blue his house, with the blue little window, and a blue Corvette, and everything is blue for him, and himself, and everybody around 'cause he ain't got nobody to listen to
@@Knowbody42da ba dee daba da.
I absolutely loved the garlic talk. Had me cracking up. My anecdotal evidence: my father (and I) always ate copious amounts of garlic growing up and essentially didn't need bug spray to keep the ravenous mosquitoes and black flies of Maine away while everyone else got eaten alive.
Thinking back, I think the sickness resistance was also there for me. Marrying my wife cut down my garlic intake and I've definitely gotten sick more than I used to, but I've also gotten old (42 now) and have three kids including a 2 year old in daycare, so who knows. Time to up my garlic and see what happens 🧄🧄🧄
I also eat a lot of garlic, but mosquitoes annihilate me!
Garlic has vitamin C which helps your immune system function. As well as magnesium and niacin, with have a massive variety of benefits for your whole body.
Before my 40s, I, too, rarely got sick and my symptoms typically lasted only a few days. In January, though, I had cold symptoms that just would not go away, even after several weeks. I finally bought a pack of Airborne immune boosting tablets, which are basically just high doses of vitamin C, and my symptoms went away within a day. I concluded that I must've been vitamin C deficient, so I've been eating a couple of vitamin C gummies per day since and haven't been sick again. So, instead of garlic, you might want to just try taking a vitamin C supplement every day and see if that's enough to get your immune system back to how you remember it.
Kids in daycare = constant virus exposure
Your wife is very happy right now. 😆
I solved it.
Garlic makes Steve's breath so bad nobody gets close and he gets to avoid all the pathogens.
Was that how garlic worked all along?
@@JohnSmith-ux3tt yes
I work for a huge company in the UK. We have a large amount of IT equipment (laptops, desktops, phones, tablets, etc). The companies policy is such that unless the item automatically bios updates then the device never has a bios update. We have a large IT department located in another country who are unaware of the Intel problem. At what stage is the mainstream public going to be informed? I only know because I watch This channel, Steves channel, etc not because of news items on mainstream news platforms
26:25 I think it’s the Australian accent, other people hear us putting an R on the end of random words when we don’t hear it. There are videos about the Aussie phantom R with examples 😂
Its a filler while the person is thinking. Alternatively people may say "um". Can be a sign of nervousness or just trying to speak to fast.
For example, 29:01 "idears"
@@IndellableHatesHandles What do you call a deer with no eyes?
When you say "other people", do you mean predominantly Americans? Because I'm neither American or Australian and I don't hear it either.
@@Woodzta Was gonna say. There's a lot of American zoomers saying "naur" or "noor" all of the time and I had no idea what they were talking about
44:21 "all customers will die within 3-4 years"
That would get Intel into even bigger trouble.
Seeing the data coming from Wendell, Level1 Tech, combined with other reports, it is established that failure is always associated with high demand communications with connected devices: RAM, NVMe, GPU - and Wendell even showed that using four sticks of RAM was far worse than using two.
So it seems obvious that either the Ring Bus is being overdriven, or the IO is overloaded. From 12th Gen. to 13/14, the only real design change was to increase Core Cache size. So it appears that they're asking too much of one or both of those data channels.
Assuming they aren't lying about manufacturing defects.
Do you think with Gen 15, Intel addressed this possible issue? I really am ready to build a new pc after 10 years and Arrow Lake looks really appealing.
@@diamondheat9 all depends if it's a complete architecture rework or re-hash. but no one will really know until they're in the wild.
@@sirmonkey1985 well it should be a new architecture since it's using a smaller process Node and Gen 15 is using chiplets. I hope someone at Intel has the same though process as you and overhauled the Ring Bus.
@@diamondheat9 Just get an AMD CPU at this point
Buildzoid did a video about the ring bus being a likely culprit a week or two ago.
AMD probably saw the Intel shenanigans and thought: oh shit, we gotta double check this.
Funny thing is they don't have to. They could literally put out CPUs that would blow up in a week and people will pretend like it never happened.
Remember what happened with ryzen 7000? CPUs and motherboards burning up? Yeah everyone forgot about it.
Come back in 5 years, people will still remember that a handful of intel CPUs are crashing using the stock boost algorithms.
Whoever makes the cheapest cpus will always have the hand up, at this moment its AMD. People will fabricate any excuse and put that as the reason why they are not buying the 14900k when in reality its just that they are literally broke af.
@@chakcrak933 except that was motherboard vendors going outside of spec and AMD quickly offered a solution, which isn't like the Intel situation at all
@@chakcrak933 Everyone has their own reasons not to buy 14900K. If any of my friends were going to buy that cpu, I would whip that friends ass. Dead platform with problems, and new platform coming later this year. Use your common sense.
@@ChristmasCrustacean1 Intel *blamed* motherboard vendors, and there may be some truth to that, but as more details come out it appears that the motherboard "overpower" issues were an exacerbating factor at best; the chips still allowed themselves to operate in ways that were detrimental to themselves.
Ok intel shill @@chakcrak933
Steve still getting 60fps at idle.
"it's only affecting some CPUs"
Well, I guess 50%+ is "some"
I had a 13900kf for about a year now. With stock settings turning off all the limits. It reaches 96c on a custom water loop. And multiple cores throttle. The newest motherboard bios for asus that ensures intel standards did not fix this. The only fix for this was undervolting the cpu - 0.0700 volts. Now no more throttling and a higher cinebench score. max temp is now around 86c max. But i still get a issue at startup where the cpu spikes to 1.5v. These issues are ridiculous and should be fixed asap.
The problem for Intel is that even if they replace 13th and 14th gen cpus with 15th gen cpus without the faults, the owners are still out a motherboard.
People need to file FTC complaints and state AGs if they are affected by this. Government has been functional under biden, so the ftc has been going after businesses cheating customers. They need complaints to act.
There is no reason they cannot use their stock of the same chips, update them, and send them as replacements. (when the microcode is ready)
thats not happing
Bartlett p core only variant on the same socket is being planned … would not mind a swap
even if they do replace the cpus, it wont be with a 15th gen...
Intel should refund the mobos together with CPUs, seeing as how that mobo will be practically useless without it, and there's likely not going to be anything to replace the chip with. Of course, they won't, and it also will do anything to weasel out of replacing the chips as well.
Pretty sure that Intel knows why the CPU's were degrading the issue is that they don't wanna deal with the already degraded CPU's cause apparently they have been denying RMA on a lot of them before Wendell and Steve released their videos so if you have an already un-stable CPU what ever fix Intel is releasing will likely not be fixing your issues the fix that is coming is to stop the degradation and not a fix for those CPU's that have already degraded.
It seems 100% intentional to me that they wanted to sweep this under the rug and let consumers deal with the cost of their stuff up
Most likely, not stop, but slow down degradation.
And it's been going on at least a year. This can be company wrecking.
The lack of real journalism in the tech industry is the only reason these companies do not feel the need to respond. They have to be forced to make changes, WE ARE THE CHANGE.
Tech journalism has never ever been as good as it is today, not even close.
@@TheHardwareUnboxedPodcast I agree, I just think the bar is really low because of conflict of interests...I bought a gamers nexus hoody the same hour the Asus video dropped to support them, would do the same for yall too. I have in the past supported both of you via patreon.
@@marktackman2886 Gamer's Nexus, Hardware Unboxed, and Level1 are the the best of breed for accurate information without BS speculation and nonsense. I won't even bother with LTT anymore because I think they have, or are prone to having, conflicts of interest.
Yeah, telling all my friends to avoid Intel for now, just an untrustworthy company.
There's a lack of real journalism period. Most outlets are de-facto propaganda outlets be it for governments or corporations.
My parents and grand parents always told me garlic is good for the immune system and when sick
Your parents and grandparents were quite right!!! 🧄🧄🧄
(Removed details about my inept first attempt at undervolting, lol.) Games have quit hitching, so that's ...nice. Cool stuff, Intel. Mad respect to AMD for stepping in and preventing an Intel-scale CPU debacle. Great PR move for AMD, too. AMD does what Intel isn't Intelligent enough to do. Doesn't quite have the ring of Sega's slogan.
That didn't helped my unstable 14900K, what it turned it back to stability were, lowering down the voltage to 0.85v and capping freq to 5.5ghz. I still have PL1 at 253W and PL2 to 288W with ICC 400A, and now it is pretty rock solid even if it ended as a nerfed down 13900T
@@Pacho18 nice. I tried undervolting further than -.1v and the thing wouldn't even boot. ETA: I was doing it wrong, lol
@@justdointhisforthegames certainly i just disabled all thermal and voltage optimizations on my msi x790 pro and it dropped down automagically from 1.2 to 0.85v, not modified any offset or did manual undervolting
@@Pacho18 neat trick. i might try something like that. brb while i brick my pc, haha; glad it worked for you but i don't trust my motherboard that much
I am 13900k limited to 262 watts, 1.28v undervolt, load line level 6, cores pinned at 5.5/4.3 on p and e cores. And not 5.8 ghz boosts. Been stable since beginning, not hitching. First chip degraded even with low VID voltages, but first cpu was limited at 300 watts. With Multicore Enhancement ON it will draw 350 watts or so. With it off it hovers around 255 watts or so in benchmarks. 2nd cpu limited to 262 watts is fine. You doing 125 watts is beyond hilarious. You might as well stick 10900k at those downclocks on your cpu at that point and 10900k probably will win.
A minor correction to your statements around the 59min mark, where you said that AMD was nearly bankrupt because their processers weren't very good. You're missing an important detail, in that the years prior to that saw Intel literally paying vendors NOT to sell AMD products. There was definitely some mismanagement on the part of AMD leadership (like acquiring both a fab and ATI when their budget definitely didn't allow for that), but the driving force was Intel blocking them from selling their products to OEM manufacturers. This led to the decline in their CPU performance, which led to the David/Goliath situation we saw between AMD/Intel.
Australian pronunciation of BIOS sounds like Bias pronunciation in Murica. In Murica BIOS is pronounced Bye-OH-s
Steve is gonna need a small auto-translate to pop up every time he says bios from now on :D
He's not saying BIOS, he's saying "Buy Aus". He's just doing his part as a proud kangaroo wrangler to support buying Australian goods.
UntiL AMD demonstrates the same level of systematic sleaziness seen with Intel, I'm not going to place them in the same group. The current lawsuit-friendly shady behavior by Intel regarding this serious issue is in a league of its own.
I like the freedom to destroy things I buy through my own stupidity. Manufacturing screwups or out of control parameters I’m unaware of is an issue.
I’ve got a 7950X3D I’m all for a Bclk overclock to get better than 5.25 on cache cores. As long as you know all your voltages are safe running delidded w/custom loop like I am, crank it. I
Blaming Board Partners ignores the fact that ALL of Wendell's data comes from Server boards that have ZERO overclocking and are configured to the minimum specs for these CPUs.
someone said they were running 6ghz 1.5V around 65w. because the server was running like 2 cores so the boost algorythm kicks in. and btw it boosts that high without any extra overclocking
tldr 300w 1.30V dosnt degrade. while light loads of 65w 1.50V do
You got the bullet the micro code are from them not partners🎉.
@@LeafGreenHDDwhy is that? Voltage is the important factor here?
If I was the board partners I would be steaming at this point. Intel better be careful or they will have their own class action.
I would not say everyone has to keep eating garlic, but it indeed contains allicin - a strong antibiotic. I do eat garlic if I get a cold.
But if you get a cold that is a virus, not bacteria 🤦♂️
Allicin is irreversably deactivated at pHs lower than 3, meaning you aint getting shit out of any garlic you're eating.
Allicin is not super effective, but it's highly reactive, very nonspecific (meaning it works against a wide array of bacteria), and short-lived, which means lower chance that bacteria get resistance to it.
A strategy with dedicated game servers is to use a LOT of relatively low cost machines(relative being a key word). Not only does this limit the blast radius when servers go down, this strategy also keeps latency low, since you can distribute these servers all over the place. It would also be cost effective on a 'per thread' basis, assuming the chips are stable that is. But intel's issue has been a tough one in that space. i7's and i9's, 13th and 14th generation carry the plague and are dropping like flies. A lot of datacenters are buying up 12th gen cpus, but those aren't stocked anymore, so are being forced to pivot to Ryzen cpus.
You don’t have to update BIOS to update microcode. The OS also contains copies of the latest microcode and applies it as the kernel starts.
Correct it is applied to the cpu when it boots into special patch ram for microcode
I genuinely can't believe people were still buying Intel. They were MORE EXPENSIVE while providing less performance (on top of higher power usage).
I guess brand loyality is one hell of a drug
You are doing the same what you accuse other people of doing. If you want to jump on the bandwagon that Intel is an evil corporation and poor old AMD just wants the best for their customers...do me a favor.
Brand loyalty might have a lot to do with brand and platform familiarity. Not everybody has had the opportunity to experience both brand. Motherboard prices and features play a role as well.
When I bought my z690 and 12600k with good cheap ddr4000... 5800x3d wasn't out yet. At the time it was cheaper than the 5800X and if I wanted more features then the motherboard price was just ridiculous.
My overclocked 12600k pulls max 120w in gaming but arround 200w in cinebench R23 with 19600 multicore score. 5800X3D and 7800X3D with cheap ram annihilates my setup in gaming but on the multicore rendering ? It's safe to say AMD is better but that doesn't mean Intel cpu is worse on all front. I'm much more concerned with the 1% and 0.1% lows and yeah the X3D are the cpus for that.
Yeah my next CPU will 100% be AMD at this point, sad there isn't any competition
I think it's kind of like how people make a sports team part of their identity, even though they have no actual connection to the team at all. That team must be good or they feel less good about themselves.
They were much better (especially with reliability) many years ago and people are slow to update their mental BIOS
only way i can see comparing zen 5 to 5800x3d is that it's most probably most probable 2 generation upgrade cadence, i doubt 7800x3d people will be looking at 9700X or 9800x3d, at least not majority
I gen skip so 7000 series isn't in my future but 9000 is.
@@GoonyMclinux Zen 4 IPC compared to the gains of Zen 3 and Alder Lake is just mediocre, nice decision.
@@saricubra2867 I skip regardless, just isn't worth the 10 fps or 10% faster number crunching times.
Steve’s interim project: best bang for the buck system using latest hardware suite that provides at least 60 fps in 12 most popular games. Go!
You can do a microcode update without flashing BIOS. Linux supports loading new microcode before/after booting, Windows should support the same.
Never doubt the efficacy of the placebo.
There's a leaked photo showing a Typo on the Heat Spreader of a 9700x.
Basically it seems like a bunch of Ryzen 5 and Ryzen 7 CPUs were mistakenly labeled as Ryzen 9.
Apparently that's the QC issue causing the delay.
Intel CPU issues on the server side should have all the companies relying on them looking into sharing telemetry to discover these potentially business harming problems sooner in the future. Wendell has shown the data was there, just no one was talking to each other.
Yeah, just waiting for the other shoe to drop here. Intel's problems are certainly much bigger, how much remains to be seen. I'll be staying away from their chips for now.
My 13700 laptop started having stability issues after about 4 months. I undervolted and put kryosheets on cpu and gpu. It fixed it for now but my avg fps definitely took a hit.
Average fps are irrelevant, 1% lows is the only thing that matters.
@saricubra2867 I wouldn't say the only thing. It was quite noticeable in the game I was playing at the time (hogwarts legacy) where I went from 70's avg fps down to about 60. My 1% lows stayed about the same.
a delay is soon forgotten, suck is forever, AMD is doing the right thing by making sure that we get good chips that operate as intended, unlike Intel and what they are going through right now.
lol that intro Steve ramble. I’ve been there brother. Good on Tim allowing him the chance to unload hahaha.
Great content lads 👍
Bottom line on 13 and 14 gen Intel CPU you can NOT expect them to remain fully functional. As such I agree with Gamers Nexus, can not recommend.
So I have an I7 14700KF. Until the Microcode update am I safe with the Intel Default enabled in BIOS? Or should I also not gaming until then to not damage the CPU?
I don't have stability issues.
Thanks!
Intel Default is not the best. It sets the AC Loadline to 110 or 1.10 (depending on how the motherboard handles it), which cranks the voltage incredibly high to the point of potentially damaging the CPU.
The best thing you can do is disable the single-core boost as it is fairly meaningless on a desktop system, and only serves to win benchmarks and degrade the CPU faster. After that, just undervolt the CPU and that will ensure you get the most out of the CPU.
TLDR: Disable single-core boost and undervolt via lowering the AC_LL.
I have 13700kf just don't use your pc till New update in August comes out
@@Arclight-n1w So I have a maximum of 1.5V on the cores and maximum temp 80-85 C on CPU in gaming. It's to much voltage?
My opinion is that Intel are attempting to covid-style "flatten the curve" by reducing the rate of damage via soft limiting while they can build enough stock to stuff their, and OEM's warranty/recall supply chain.
Fact is, they simply cannot do an instant recall.
There aren't enough parts or technicians in the world to declare a no questions swap out.
With desktops and servers it's relatively easy. Build and deliver the parts, pay ~30min swap labour.
With laptops it's much more of a nightmare.
OEM's have to build replacement motherboards and fill their supply chain - and Intel has to reimburse them. Now, it might be possible to only build 20% more boards and de/resolder swapped boards, to achieve 100% replacement, but it's still an expensive proposition.
The best they can do is slow down the return rate while they implement an effective recall process with all their partners - something they haven't had to do for a long time. The institutional knowledge of the process to achieve this is long gone and/or obsolete.
The other problem is that errors are less desirable than reduced performance.
It may have reached our attention because of gaming, but gaming crashes are merely annoying in the grand scheme of things.
These processors are deployed in mission critical roles performing all sorts of tasks where errors have real economic and potentially human costs.
Financial systems cannot abide a processor that can't math accurately.
Safety and medical systems can't either.
Imagine an xray machine comparing the $currentradiation to $maxradlimit and giving the wrong result. It's remote, but someone could plausibly be irradiated.
Even a billion to 1 error will happen thousands of times a day across the deployed fleet.
"flattening the curve" is their safest option right now.
I hope you didn't vote for Intel..
Lol you guys should have put an affiliate link for the garlic. I 100% ordered some.
Dude! Garlic is very well known for improving your immune response like this. I personally hate it and avoid it, but even I one time couldn't get past a really long cold and started eating garlic in addition to taking my meds that just weren't enough. That solved it.
Atleast amd has delayed and did a recall unlike intel just letting then out in the wild
Also it was obvious in the test samples they sent to reviewers for review, that they had a issue/s, but they haven't revealed what it is yet as of 2 days ago now. Speculation wont help unfortunately here.
Yeah, lost all trust in Intel
i may sound like an AMD fanboy, but i feel like there is just more good will in AMD, all those open source projects, all the things you guys mentioned, 7800x3d melting issue, they did cover for Asus error, while year later Intel throws all board partners under the bus
AMD does have a wierd marketing team, which quite often makes things awkward, but they are company, profits is all that matters, but from my point of view, in the end one thing matters for a consumer, how they solve the problem
for example, Nvidia blocking FG behind 4000 series for artificial reasons, AMD could do the same for AM4, just say that zen 3 needs some hardware tuning in chipsets, so only 500 chipsets can work properly and push that, some will say they crumbled under the pressure, but in the end they just did the good thing, same with FSR, it being open source gave a lot for GTX and RTX users who wants Frame Gen or plain old upscaler, Intel also got a lot of praise for XeSS being open
Digital Foundry made a clip about AMD research about hardware RT and Meshlet technology and both of them are not locked by any hardware like Tensor Cores, so i think that kind of openess gives the feeling of being trustworthy, but in the end, yes they are a for profit company and have share holders to satisfy
The best way not to be a fanboy is to recognize the most likely reason behind all of this: they are not the market leader in any of those categories.
@@roboninja1 not in GPU for sure, but in enthusiast desktop CPU and servers, i believe they are, better or same performance, more power efficient, no need for industrial grade cooling solution, cheaper, 7950X was what $100 cheaper than 13900K/14900K? Intel strong side is mixed workload gaming and professional on i7 parts, other than that it's mostly within 5% difference, so kind of nothing soup, you'll be happy from either of those as long as they are stable
thing for system integrators is like someone said "you don't get fired for going Intel" they were golden standard for so long, that it was just safe bet, but now with all those problems it will stir the pot a little
and locking features in doesn't take much, it's even easier since there's no need to make it accessible as much
The Intels in my past: 8086, 386, Pentium, Pentium MMX, Pentium III, 6700K. The AMDs in my past: 286, 386(40), 586(133), K6-III, Athlon MP, Opteron, Phenom II, Bulldozer (Vishera), and now AM4. The fact that Intel started requiring new motherboards each year ended it for me. I have an awesome motherboard for my 6700K that should have been sufficient until at least 9th gen. My rule of thumb is if the PCIe generation doesn't change then you shouldn't require an upgrade.
@@1Grainer1Raptor Lake mobile is lightyears better than desktop.
The Core i9-13980HX is extremely impressive, it's faster than a desktop 13700K and more power efficient.
The perfomance per watt is also better than a 13900K/14900K which is funny because it's the exact same sillicon.
A friend has a Ryzen 9 5900X desktop tower and he uses that i9-13980HX Asus Strix laptop for work during travels.
Crazy perfomance in Baldur's Gate 3's city (which is extremely CPU intensive) and i didn't hear any fan noise.
@@1Grainer1 Huh? There has _never_ been a point in time that the 7950x was cheaper than the 13900k. 13900k launched at $590, 7950x launched at $700. 12900 launched at $590, 5950x launched at an eyewatering $800. 10900k launched at $490, 3950x launched at $750. Sometimes the price gap is big enough to straight up buy a new motherboard. Even right now the 13900k is $440 while the 7950x is $525.
You can complain about efficiency and thermals and that's more than fair, but Intel is flat out _NOT_ more expensive than AMD. AMD has been going for your wallet for years now. Anyone that claims otherwise is a straight up fanboy spewing copium, or simply doesn't know what these parts _really_ sell for. The fact you magically think the Zen3 on non-500 boards wasn't an issue is pretty telling. Especially when 400 owners had to wait 4-6 _months_ for BIOS after launch, and 300 owners had to wait a laughable *_18 months?!_* That's not doing the "good thing". That's a straight up backhanded tantrum that they were pressured into offering compatible BIOS. That's the kind of "compatibility" Apple gives when the EU sues them for anti-consumer practices.
_EVERY_ time AMD has been in the lead or even matching the competition they've charged sky high prices. They did it with Athlon, they did it with FX and their 8 cores. (The ones they got sued for?) They've done it with Ryzen. And how's that Threadripper pricing today? AMD is NOT your friend and NOT the "good guy" as much as you think they are. They do it with CPUs, motherboard chipsets, as well as GPUs. (Remember Fury pricing? Vega pricing?)
@roboninja1 absolutely has a point that they only get honest when they have competition. _Exactly_ like Intel and Nvidia. That's not to say Intel and Nvidia aren't just as shady, but to pretend AMD is somehow any more noble is just plain naive. I daily drive AMD and Intel systems and I hate ALL these companies. But it's a straight fact that my one AMD system has had FAR more issues than the 3 Intel systems in the house combined, and it wasn't the cheapest one either. And yes, one of those is 14th gen. I would definitely say your fanboy glasses might be clouding your judgement.
F Intel. F AMD. F Nvidia. None of these big evil corporations deserve your fanboism. It's brainless sh*t like that that let's these companies think they can get away with the BS they pull on a daily basis.
If the 13 and 14 CPUs are not fixable, and the replacement option would bankrupt the company, another option is to issue a credit based value to be applied to the future (next) gen CPU. This way not having to layout a cash loss with direct replacement. Regardless it is serious hit on the company not just now but going forward as well.
Hi guys. I’m currently shopping for a work laptop upgrade. Would you suggest i stick with the i7 14th gen instead of the i9? I’m purchasing in the coming two weeks as I’m still researching.
I'm still on the fence : am i in deep sh*t with my core i5 13600k? I own that CPU since january 2023. It is installed on a Gigabyte Z790 Aorus Elite AX DDR4 mobo. My casing has plenty of airflow (EATX Fractal Torrent) and my cooler is the great Noctua NH-D15S Chromax Black. The fan curves are fixed in a way that my games rise the CPU heat at 69-74 , max. It's not overclocked and I dont know how to undervolt it If it's required to give it a better lifespan. I watched some videos out there but none of them were clear enough for the casual users like me. I need a clear newbie guide step by step. I downloaded the bios update, extracted it at the USB key root. When i boot my PC with it, the boot logo show on screen but nothing else ever happen.
The fist time you said bios was strange... but nothing beats TIMs HGR (HDR)..
I want to know more about that Wizard and where to find the Wizard... I have back issues and it just keeps coming back every 2 months.
Garlic is a known antibacterial agent. I would love to learn other medical treatments from Steve's Wizard. :)
I do QA at a software company, at first our clients had full access to the whole software, then we had a new CEO and he decided we need to put limitations so that the clients pay more to have those limitations removed, which is a practice I really hate, so anyways, he said it so we had to implement it, then while doing QA I noticed that there's a way to bypass the limitations, I validated the change and told nobody about that, although I was almost sure none of our clients would know how to bypass the limitations. Sadly, 3 months later another dev discovered the bypass and patched it.
Right now they want to prohibit account sharing to force the clients to pay more, I am against it but there is nothing I can do about it. Also it's pretty hypocritical because we do share accounts on some services we use.
So yeah anti consumer decisions do not reflect the employee's mindset, but rather the top level management's
They get dollar signs in their eyes and don't realise they are destroying the company's reputation.
Hey Steve! Which specific garlic product are you using?
My more-than-a-decade-old Intel Core i3-3220 is still kicking around just fine in my old PC, and here I am, worried about my 4 months old 13600k degrading over the next couple of years, I was planning to use to for at least as long as my previous PC but it seems like it won't even last half as long as that one. If Intel can't fix this problem, then Intel should be sued to the ground for this unless they provide us with refunds and compensation for buying into their platform/motherboard that only works with these defective CPUs (technically it'll work fine with 12th gen but that's a performance regression so it shouldn't count).
The notebook cpu issue it's underrated. A desktop cpu fails, you can at least replace it, at most together with a mobo. It also has a 3 year warranty. A notebook CPU dies, many notebooks have warranties of 1 year, 2 years tops, and mobo+cpu replacement, even if available, is usually as expensive as buying a new notebook
The garlic story actually make sense somewhat. In Greek culture garlic is believed to grant eternal youth or something like that, it definitely has some positive health related reputation in tradition. Thats why the Greek put copius amounts of it in food. I'm not even kidding
55:19 they definitely are becoming more anti-consumer. I wouldn't be surprised if B840 is just a first step in increasing the average sale price by slowly eroding the B-series until it becomes a hollow husk with none of the features that made B-series so good.
I bought a 13700K in late 2022, so it seems likely mine has the oxidation issue. If that's the case, there is no fix other than replacement. But so far I've had no issues. Important question: what's Intel going to do for customers who have this ticking time bomb that hasn't exploded -- yet?
absolutely nothing; besides stalling until your warranty is gone.
@@572089 I guess I'd better start having issues then.
Does the 9th gen i9 9900k have the same issues of degradation? I can not run mine on turbo with hyperthreading. Can't clock memory to 2666 or system freeze will happen in windows.
You get a lot of vitamins and minerals from garlic. If you wife can't eat garlic, I'd recommend golden kiwi fruit, as they're high in Vitamin C, and are much nicer than regular kiwis. Or any citrus fruit really. Just avoid replacing fruit with fruit drinks, as they often lack the vitamin C, as it breaks down in the production process, and is often flavored with massive amounts of refined sugar, which is obviously bad for you.
It also contains a not very effective but short-lived and nonspecific antibiotic allicin.
I heard Garlic is a natural antibiotic so maybe there is something to it
I can’t begin to imagine the average user opening their pc and replacing the CPU after removing the attached cooler. So not only is this a part replacement issue but a huge labor and shipping cost.
Buildzoid had an interesting video also referencing Igor’s lab test on cpu voltage and it seems due to binning that certain CPUs are basically running stock over voltage and others have much lower voltage. Some were stock almost 1.5V and others stock 1.3 and less. This on both i5,i7 and i9. So your mileage may definitely vary. But if your cpu is stock running close to 1.5V I assume that is definitely going to impact longevity.
I am OK with the delay. I have a 7700X in my SFF build and I am not looking to upgrade at this time.
I think AMD is doing the right thing by pushing out the delay to make the launch successful.
Steve does his best Wario impression in this one.
Steve interim project suggestion; Sell me on DeWalt. I've been a Makita man for a while. Had a bad experience with a DeWalt drill (corded) so I've been a little nervous about them. Give us more some Hardware Toolboxed.
would be nice for affected EU intel consumers to write the EU body tonforce a recall
I think the reason for AMD delay will be visible on release notes for new BIOS releases for motherboards that already released their Zen5 BIOS-es prior to launch. That is if something had to be updated via microcode update.
I really hope it doesnt affect the 13900HX (mobile cpu) where it is a lot more difficult to replace. Lenovo recently replaced my Legion 7i Pro motherboard due to an issue where the symptom was the CPU reaching 90-100 degC at idle. This issue popped up after less than 12 months of ownership. While there are other potential explanations for this, it was really odd due to how it came about. Fingers, toes and doodle crossed that this insidious problem isn't the cause as it is a 4 grand laptop.😢
They give raw garlic too to cows to not pump them full of antibiotics is a good remedy an not snake oil
Btw you find over the counter raw garlic in flexible shells so yo can skip the bad breath and all
meanwhile on intels side "Sir we can not compete with Ryzen" and the boss so "Shut up and sell them for double the price", Worker "But sir the Products haveing issues" and the boss "So shut up and increase the price again so we get the most amount of this crap".
AMD was (still is?) in similar place with the Zen3 WHEA error 19 debacle, and to some extent with the X570 USB disconnect problems. I had to return my 5950x which was not stable in light threaded workloads out of the box, and I was far from alone in this boat.
You can test the 14901KE. You’re fine on BIOS, but in US English, the O is pronounced like “Joe” or “Row”, as how you all pronounced it secondarily.
To add some further mysticism to the wizardry, I would add a past gf who was Chinese very much believed in traditional medicine and had me drink ginger garlic tea when I would get sick and it helped me recover every time. Garlic has anti-microbial properties. You can look into it more to get the proper explanation so Steve's story makes sense to me. We ate a lot of garlic in general for that reason. Chew on some raw garlic for a sore throat if you can handle it
seeing how AMD handled the x3D chip failures i'd like to think they'd handle it better but agree it's hard to say exactly what they'd do given how much of the customer base it effects(OEM/server/DIY) and wouldn't surprise me at all if they ended up doing exactly what Intel's doing right now. but i also think the way AMD handles the bios limits and restricting board partners on what they can and can't change is done a little better.
For people using 13th/14th gen cpu's what would you recommend they do/check to see if their cpu is degraded or having issues but not dead yet... I do not expect Intel to really help end users out
I think it'd be funny if it was just that they mixed up the bins and were putting like 6 core in 16 core boxes and 16 core in 6 core boxes.
"optimistic" binning assumption is reasonable since their chip making division is running a net loss, so they're stimulated to sell more high-margin SKUs
My TUF Z690 board has a new BIOS released this week from ASUS, stating to have the 0x125 microcode fix. I've got a 12700KF so I don't know if I "need" this update.
The 12700KF is so power efficient that at stock it uses low wattages and voltages.
It's literally better than the entirety of desktop 13th gen and 14th gen K chips.
All 3D V-Cache CPU's vs their non-3D V-Cache counterparts in gaming and productivity. Make it happen. 😉
I have a 13900k with 2 sticks ram (xmp on) at 6400. CPU cooler is 360 Deepcool AIO. My system was made around late December 2022 haven't had any crashes OR bluescreens. Am I good how do I check if I have degradation?
I have a 13900k since 29/12/2022 no crash and no stability issue. Maybe all CPU are not impacted
Editor making the thumbnail deserves a tenner extra.
...could (some of) the board partners actually sue for defamation/fraud over Intel's assignment of blame to them and their power settings?
Buildzoid made an interesting observation about e-core frequency boosting causes the chips voltage to go over 1.6v. I'd try disabling the e-core boost and see what the effect is, and what type of performance differential it would create.
Is it not illegal for Intel to refuse you a RMA if you refuse to “down clock” your cpu? You infact paid for advertised speed and a down clock is below the speed you paid for?
It doesn't matter because corporations in America are above the law and the government only enforces the laws on the plebs.
Came for the tech news, left with a desire to drink liquid garlic. Thanks guys!
I bought a $380 motherboard to OC my 13700K and now I can't. I'm furious! Intel can go to hell. I won't forget this.
It’s a huge mistake if Intel has already been selling a lot of these chips at near-zero profit margin simply to block AMD from marketshare this whole time, and this problem really shows why. If you already didn’t make money on the original sales and now need to replace parts for free, well…now you’re shedding money for no reason. And then speaking to what’s being articulated roundabout midway through, if you RMA the hardware without having a software resolution to adequately protect the replacement hardware you’re just heading for another round of RMAs down the road.
Intel’s hope definitely has to be hoping this can largely blow over after ArrowLake releases to minimize RMAs they have to honor, and they have to accept doing ArrowLake properly beforehand this time even if they have to cede the performance crown another generation.
There are so many levels on which this entire problem has been mismanaged and it’s really putting Intel in the hotseat. Intel needs to acknowledge internally they aren’t in a position to continue pretending they’re in the lead anymore and simply compete where they actually are.
Buy these HUB Nutriceuticals!
"Free from Oxidation and runs your Intel processor up to 20% faster"
24:32 Yes, many people say bye-ose. Most, by my reckoning, but I haven't done a survey. Saying bye-oss sounds strange to me and many others, so that may be what's going on here. Probably just a regional pronunciation thing.
According to Moore's Law is Dead latest video, the oxidization issue occurred in their Chandler AZ Fab between March and June 2023. We still do not know if those CPUs made it out to the channel or batch numbers that the end user could check to see if their 13th/14th CPU are from these batches.
Kinda interesting to hear handball referred to as obscure - but I suppose for Americans and Australians it is. Oddly enough, I think it could be very well received in both the Americas and Australia, especially with some minor adjustments to rules and aesthetics. I imagine it's more of a coincidence than anything else that it isn't a thing in those areas.
Honestly my theory is that the oxidation issue is the main issue but since that is completely un fixable for existing cpu owners they throw the voltage excuse out there as a possible xause as a least worst explanation that you can get a microcode fix for.
Then when cpus fail anyway, people just assume it was too late that the microcode fix got applied and they were just unlucky.
It isn't as if every cpu owner has a materials analysis failure lab to call bullcrap.
The fact that they said the oxidation issue was only for a period of time and got fixed back in 2023 but they have not posted a website with all the affected cpu serial numbers for recall for an apparently known date range of affected cpus. That is highly suspicious that the issue was actually for much longer than they are letting on.
A live version of these would be interesting.
It's the same thing :)
@@TheHardwareUnboxedPodcast Fair enough.
Someone needs to get this issue on the big, mainstream news channels. Imagine listening to those pundits talk about "Intel CPU Instability" and "Microcode patches", and basically everything in the Verge article published just before noon yesterday.
That would light some fires.
ive never been so aprehensive for buying new gear than now.
Yeah, crazy how Intel is the unreliable one now. How times have changed
They are joking about the garlic shit but garlic is seriously a wonder herb