Is the general consensus that a delay seems better than a botched launch? I'm sure AMD was looking at Intel's situation and considering the optics of their own problems. Help support our independent reporting! Grab a PC building anti-static Modmat here: store.gamersnexus.net/products/large-modmat-gn15-anniversary Or one of our project & soldering mats! store.gamersnexus.net/products/gn-project-soldering-mat Learn more about the Intel situation here: ua-cam.com/video/OVdmK1UGzGs/v-deo.html Read our ad-free article! gamersnexus.net/cpus-news/amd-delays-ryzen-9000-did-not-meet-quality-expectations (support via the store & Patreon!)
Yeah, my immediate reaction was “at least they caught it before they shipped to customers”. I’m not too bothered as long it’s not customers stuck with the headache
Yes, Delay something until its done right. At the price of $600, people expect the product to meet expectations and work correctly, if it needs another month, or even longer. They should take it. Much better than selling chips that can't handle such expectations.
A delay of two weeks is infinitely more preferable to nebulous issues down the line. No one will remember this in a few months. Commence unclutching pearls. EDIT: to be clear, I was not suggesting Gamers Nexus were the ones clutching pearls.
But what can they possibly fix in such a short amount of time thought? Can't be a hardware issue, possibly just a software/bio fix or something, interesting nonetheless.
This is almost too good for AMD. By holding off and doubling down on production performance perfection as a poke at Intel, is straight up publicity gold for them!
@@SimBunkerThey aren't. If they were, that would mean they were already launched (when launch was supposed to be end of this month) and then they couldn't be delayed, could they? If by saying "in stores" you actually mean "in warehouses": That still means they aren't in stores and can thus be delayed. What was your train of thought there, mate?
@@Behemoth33 everybody? That's how these things always go, security patch, tank performance. Stability patch? tank performance. The reason it's unstable and degrades is that it's pushing too much for performance
Absolutely fine with a delay over quality issues. That actually makes me more excited for the new chips, since AMD at least appears to be focused on delivering a quality product.
@@fajaradi1223 I think it's most likely they want media outlets to test the new Ryzen against Intel **after** the patch, which in all likelyhood will decrease performance. This is Intel trying to squeeze by and AMD saying "denied."
''Delays launch to meet quality expectations'' That's exactly what videogame companies say when they delay a game by two weeks. Those two weeks never change anything.
@WrexBF Of course it makes a difference. That difference being a pile of files that insta BSOD's, and something that resembles a somewhat functioning piece of software that may be a game someday lol.
@@WrexBF how do you know this? Have you gone round and played a good number of games at their planned release date and then their actual release date and compared them?
I like it also, but it's not like amd nor intel would answer truthfully to this question anyway. It's in their best interest to not ship golden samples cuz it'll quickly be found out. But it has happened in the past, so it's a valid question. But i doubt either side would answer truthfully on that question. But ye since they're recalling a whole lot more than review samples, i guess i could believe them somewhat on this one.
I’m confused, why didn’t they just increase the voltage of early chips to degrading levels so they could hit performance/stability targets? That’s what Intel does and that’s been going well for them!
Because if AMD did it all the Intel fan boys would wet there pants 😂 also AMD knows what happens if you fluff a cpu (Bulldozer and piledriver anyone) nearly went bust over it and lost a lot of good will, am 99% sure intel legal bills will put them at rist of going under as its going to be billions in liabilities esp for servers and OEMs Dell and HP ech are not going to want to recover there costs for sending engines out to companys to replace cpus as well as reputation damages
@@kramnull8962this one was actually mobo fault, and even then they just acknowledged it, refuded or replaced the chips and done. Sometimes things go wrong and it happens, it's how companies react to it that truly matters.
As someone that had to do a RMA of a 1800x due a memory bug that affected some of the inicial production units, finding this before retail sells them is a good thing.
Zen 1 was such a rough time not just because of all the issues but because absolutely nobody had any experience with the platform so most advice you received for issues was like alternate medicine tier placebo effect stuff lmao.
@@rain8478 Zen 1 was rough in general for its whole lifespan. Mfgs didn't take Ryzen really seriously until the 3000 series. The motherboards were lacking so many features compared to Intel that you really gave up a lot of stuff for those extra cores and threads.
@@rain8478 i’m at the same boat. 1800x -> 3900x -> 5800x3d, happy with the decision at the time and don’t feel the need for a plataform upgrade any time soon.
You know what? On the one hand I'm super disappointed because I'm overdue a CPU upgrade, but on the other hand, I would rather that companies release products when they're ready as opposed to when they look best on a balance sheet.
Even software companies that used to "release when it's ready" are under pressure and time crunch to release on a schedule. Probably hoping to fix on a XX.1 or z-stream release.
@@junko4166 I'm currently rocking a 5900X with 32GB of RAM that refuses to clock above 2433 (in spite of the advertised 3800mhz), and I'm doing a lot of AVX workloads where Ryzen 9000 should be a monster comparatively, and I want to upgrade to 192GB. It's not really for gaming so I actually do want to upgrade very soon, and I honestly need the RAM, too.
Imagining the logistics involved in pulling products back, evalute and ship out again, 15 days sounds like an impressive turnaround time. AMD must be super confident it's just the first batch.
They don't evaluate and ship them out again in those 15 days, they ship out completeley new CPUs that were already evaluated correctly. The old CPUs then can be re-evaluated at any given time
@@TheTaxxor True, obviously they could supply replacements even before receiving a shipment back. Yet, with linked reputational risks I'd expect they need to confirm via some re-evaluation before giving the go ahead. Probably they first retest trays, rather than ripping boxes open..
"A delayed game is bad, a rushed game is bad, all video games are bad, I fucking hate video games and making them." - Shigeru Miyamoto, gaming industry legend
Ha Ha the prices have just gone up as nobody will buy Intel right now! I hope Intel get through it as one CPU manufacture would be boring. Im in bed with both on my current systems but my gen 13 in mobile 😅
Just makes it look like they don't know what the hell they are doing. The reality is overall AMD has had far more problems with cpus and compatibility issues than Intel has had in the last 6-7 years.
Fr, i still feel like it hasnt been that long since ryzen 5000 was released We are so used to everything moving so fast in this industry, it starts to look really absurd when you step outside it for a while and come back
@@pituguli5816 No one is going to have the CPUs regardless, so what does it change for reviewers exactly? Plus, I don't think AMD really cares if a few dozen reviewers are getting inconvenienced compared to the millions of customers.
LMAO, the INSTANT Intel released their announcement, my first thought was "AMD should totally delay their release just to piss off Intel". Good to know I'm in tune with their management
On one hand I would love for AMD to be shrewd enough to actually think like that, but on the other hand I'm also happy that they are not. They've always been "dorks" at marketing. However fun the thought is, I think there is a legit concern of something here. That we haven't seen pictures of Ryzen 9000 boxes at retailers yet, and the like, suggest that they have been on the fence about this one for a while.
I do validation for automotive chips and the timeline doesn’t surprise me too much. I’d speculate that there’s a process parameter that’s causing issues and wasn’t being reliably caught by the testers. So when validation does the last bulk unit screen on production parts, before launch, some units pop up bad and they have to raise a red flag. The failure rate also needs to be high enough that C-suite doesn’t think they can brush it under a rug. Chances are they updated the testers to catch the issue and affected parts will be binned down.
@@TheLK641 whoops, yeah I meant the ATE systems. Figured most people wouldn’t know what those are but probably should have been a little clearer than “testers” lol.
Don't worry, it made sense once I reread it, I just find the idea of a company going around updating their employees' firmware very funny for some reason *-*
The ad for the AIO at the beginning of the video is so authentically honest and positive, with pertinent information about the quality of the main fundamentals of the product that one would be concerned with, and it may be the best UA-cam sponsor spot I've seen as far as making me want a product based on the presentation
i actually clicked on the link in the description just to see what the price was. $60. If I end up needing an AIO anytime soon I guess they'll be first in line.
@@SLLabsKamilion i got the 36o rad from them when it first came out... after a youtuber said best new hone brand..its almost seems too good to true...the business model is they are taking noctua tech from their last gen and making it cheap.
Seems like GN gets to choose their own words, with the sponsor just taking a brief glimse over it and giving a thumbs up. I am not interested, but it's good presentation.
@@edwardallenthree If it is a marketing ploy, it's a fair one. Even if they're just pulling the tail on Intel, for obvious reasons... technically all of these manufacturers should be taking this approach of double checking there are no issues before selling their product anyways.
@@Lookingformorefunyou might if your computer is your brain rot editing and mass content farming device, keeping you off the streets in the absence of any marketable skill. otherwise; you’re correct.
9:36 Agreed. Especially with the community sensitized to issues, AMD needs a flawless launch to demonstrate superiority. Heck, this announcement is probably a fantastic PR move!
@@erkinalp I am totally Team RED, and it's a decision I'm quite proud of. AMD is a great company, led by engineers, as you point out. That's a huge difference!
Delayed is always better than poor quality at launch. We are learning to prefer this in games software, so honestly glad to see AMD focusing on quality in hardware land.
Learning? We always prefered it that way, the internet and with it the ability to fix stuff later gave devs simply a tool to push stuff out before its cooked to perfection and maybe fix it after cashing in.
Sometimes delaying something removes the first-to-market advantage and it may not be worth issuing the product at all. However, I think AMD is smarter than that, especially immediately after Intel has ventilated its own foot.
@@Gnarfendorf true but our tendency to pre-order everything and threaten devs over delays certainly doesn't help. The latter may be a loud minority, but it's better to make it clear that we want quality over quantity or speed here.
Oh I bet it's nothing like that, since they were claiming that TDP was down. They may have the opposite problem where they set the voltages too low and the chips crash at different power profiles from lack of voltage.
@@aegiltech What CPU cooler and motherboard are you running, I'm running the Ryzen 9 7950X3D, RTX 4090 FE on the AsRock X670E Taichi with an Arctic Liquid Freezer II 360 While gaming the 3DVCache cores consistently hit 5.25GHZ and the regular cores exceed 5.7GHZ, highest core temps I've seen is around 74C which is below the 89C TJunction Max at which point the CPU will start the throttle, it would appear your throttling due to cooling....
@@N7_Nate ASUS Extreme x670E with a ASUS 360mm Ryujinx. Was happily running at 6000mhz with 2x32GB GSkill AMPsticks (which are on QVL list) until recently. Have tried different sticks, no joy, have another system with almost identical specs running a X670E ProArt downstairs, its fine. System doesn't pass memory training at 6000mhz, or if it does, it bluescreens windows. need to try swapping CPU's with other system but have been too lazy.
Yeah it sucks that they had the issue at all, but if anything it just makes me want AMD products more because it makes me more confident that they are going to catch and fix problems before it becomes my problem
Intel has been winning for a while now. AMD needs this release to win and even if they do it’ll only be temporary as intels next chip set will again be on top Edit : AMD Fanboys jumping off buildings and the r9 is trash
i say they waiting for intel to see if they can fix it and then price accordingly if intel cant fix, that means AMD has no competitor for a few months until arrowlake comes out. they will pull an nvidia and itll be super overpriced.
Admitting part of it... They still haven't addressed 14th gen failures. Or why they sat on knowledge of this all this year so far, letting unfixed CPUs keep on oxidising
Computer hardware be difficult, Getting it right is the hardest thing in the industry. There is a MASSIVE reason why there really isnt more then just a few CPU brands.
@@tilapiadave3234 It's almost as fast as a 7600X in games, but can be run without buying a new motherboard or RAM, so a way cheaper upgrade for the millions that already own a AM4 computer. I don't regret buying mine one bit.
@@tilapiadave3234 Since it's now an older chip, if you can find it for a reasonable price it's pretty great. Especially if you're on AM4 and don't want to spend more to move to AM5.
Not double but definitely closer to re-supply, so there will be less stock drop off. Win win because there will be a lot of new customers after Intel did this dumb stuff.
Hopefully these issues (on both sides) have some silver linings for consumers. It'd be nice to see the new AMD chipsets come out at the same time as these CPUs too
Actually a smart move by AMD. They want to make sure the new Ryzen 9000 CPU models don't end up with same horrible crash issues that plagued the Intel 13th and 14th generation CPU's.
@@WrexBF You're not paying attention. AMD told that it's an issue with the first batch of CPUs that has already been resolved, so the delay is just to retrieve the already shipped units and replace it with fresh ones
You don't even have to go to Intel, AMD learns from its own mistakes. This could be similar to Zen 2 where the CPUs were not reaching the advertised frequencies by a little bit (like 50 MHz or so), that one was resolved by an AGESA update if my memory serves right.
If AMD are telling the truth, that's a good call. They're probably stalling to have the Intel microcode update out until they let reviewers have the chips =D Kind of a chad move, AMD seems to be learning.
What's this? A company that knows that it needs a bit more time to resolve any potential issues with its product? Has the bar been set so low by others that this is not even considered the norm?
@@kramnull8962 stop posting this under every single comment over here. You know very well that a) this is unrelated to this story, as it was a Mobo issues, the CPUs were fine and b) AMD did what Intel is incapable of doing: be open about the issue, fix it in a reasonable amount of time and make customers whole.
@@Hugh_IWhat you're replying to is something I see with fanboys of certain brands all the time, they reply to things constantly to spread half-truths relating to a competitor of a brand they worship.
I wonder if this is more to force Intels hand then anything wrong on AMD's end. Now Intel will be forced to push their cpu microcode patches first which will impact performance. AMD will have the luxury of benchmarking against the REAL performance of these Intel cpus and not some bogus pre-patch number.
There's no actual facts that the intel will perform worse. Remember, lower voltage = lower temps = longer/better boost But I have benched my 14th gen, will see how it will compare with the microcode results, so now that I have the basic results I'll keep that rig off and wait for the update and retest it
@@Behemoth33 If it's the actual CPU core that's failing due to high voltage, Intel will likely have to pull their boost clocks down a bit to maintain stability at lower voltages. With how high they are pushing the clocks, I doubt they will be able to guarantee a 6GHz boost with even a very slight reduction in peak voltage. The default specs they launched the CPUs with were no accident, they knew they had to push the voltage to that extreme to hit their target clock speeds, it's not like they pushed the voltage just because they felt like it.
I don't think this hypothesis holds water, because from what the video is saying their partners are genuinely sending the CPUs back and getting CPUs in return, which is incredibly expensive. GN has got third parties on the record stating that they're sending them back, which you wouldn't need to do if it wasn't a real issue. If all you wanted to do was to delay, you could bypass that cost by saying the issue was something you could patch remotely.
Damn I work at a PC retailer and have had 9700X on hand since July 6th. Getting CPUs and GPUs a week or two before launch is normal but 3 weeks seemed early so I didn't expect this! Then again it did seem strange that 2 weeks later we still only had 9700X and no other SKUs so I guess that explains it.
What a W for AMD. Even if it's primarily a PR move, it essentially draws the comparison between Intel that sold subpar/defective products to their users without saying anything, compared to a delay for quality.
Keep in mind that this A) costs very real money, and B) massively inconveniences their partners who have to deal with sending back the CPUs and waiting for the replacements. I can see an argument for how point A is mitigated by the PR win, but that's really only true if you don't majorly piss off your partners in the process. You wouldn't want to involve all the systems integrators in a PR stunt like this, and it sounds like those companies are in fact involved and sending their CPUs back.
My guess would be that SIs started doing internal testing and didn't like what they were seeing and alerted AMD. AMD did the right thing here in any case, claw back and test was a good move.
They were hoping The youtube crunch on 13th and 14th would make some idiots trash their Intel processors and run to AMD while the prices are high. Remember 100% bad cpu every one is pushing.... Amd wants to sell their AM5 backstock now, not at black friday.
My release date Ryzen 5950X suffered from occasional WHEA error reboots (not even a BSOD) shortly after launch. Later BIOS updates mostly mitigated the issue and I could go months without an issue, but over the 3 years I used it it still occasionally had stability issues. I eventually got it replaced under warranty shortly before that expired. This didn't seem to be a widespread issue, and most of the chatter about it seemed to be shortly after launch so I suspected that it only affected very early CPUs. Given the stability issues Intel is facing AMD may have caught a similar issue with Ryzen 9000 and decided that it was best to get out ahead of it.
How did you know it was a WHEA issue? Was your infinity fabric set past 1800? I have occasional instant shut offs with no errors shown in windows event viewer on same processor
@@vaels5682 Windows Event Viewer occasionally recorded a WHEA error other times it just rebooted with no error. I didn't have it overclocked besides XMP to run the RAM at 3200MT/s.
@@TCL987hmm, i'm running this sweet 3600mhz Cl14 kit (no tweaked timings just xmp). I'll have to find a good search query to cover potential WHEA errors see if i come up with anything. My mobo has dual bios so maybe ill try a bios update ive been hesitant to get and can go back to other if needed (i read the most recent bios update had slight performance reductions for a security update i didnt see as being necessary)
I've read a fair share of user reports of WHEA errors on Ryzen 5000 CPUs, but luckily RMA was completely hassle-free in all cases, which is an aspect I need to compliment AMD for.
Sometimes I think the GN videos are too long where Steve just rambles on and repeats himself, like this one could be 3 minutes shorter and lose nothing. Just a thought. Cheers, keep up the good work!
Props to AMD for making a difficult decision for the greater good. Too many companies today would have just barrelled forward, hoped for the best, screwed over a lot of customers, then played the usual PR damage control card while continuing to collect sales revenue. AMD just delaying is the right way to handle a quality issue, and it makes me trust the company more.
What a fantastic time to have to upgrade my 5800x to meet the demands of content creation. Deal with rusty molten Intel's or settle for a "random Ryzen". Pain.
I hear you. Building a new content creation computer myself soon. The 7950x is looking pretty good today. But it's probably worth the couple weeks wait, if possible. Would be nice to see if the 14th gen Intel is truly fixed also.
'Random ryzen' lmao what? 7950x is going to be the easiest jump for you. If you game, get the x3d. I've built PCs for myself, FnF etc over 20 years, my 7800x3d build was the easiest one I've ever done. I didn't even have to install drivers ffs. It downloads them automatically now O_o. Rock solid from day one, quiet, cool, fast as. EXPO DDR6000 CL30 trains in under a minute on an ASRock x670e steel legend. AMD hit it out the park with the 7000 series and 9000 series looks to continue that. Zero issues, platform is stable as and I've been watching it closely since Zen 1.
@@N4CR "random Ryzen" was an attempt at a joke, eluding to the fact that quality could not be guaranteed, hence the delayed release. I don't want the easiest jump, I want the best long-term jump. The point of me waiting this long to upgrade was to not be forced to upgrade to a 2 year old Ryzen that performs worse than a 14900k in Adobe Premiere. It was to upgrade to a substantially better CPU like the 9000 series, or the 14th gen (if it didn't rust and cook itself). But now I have to wait until they figure out the 9000 series or fix the 14th gen before the 15th gen is released (assuming they didn't use the same flawed design in the 15th gen).
I doubt it since AMD hadsimilar issues with the 7000 v-cache chips starting motherboard fires. AMD's issues were easy to fix, because TSMC makes good wafers, and it was all bios related. I think Intel has deeper problems and are pulling an ASUS move out of their asses.
Could just be a strategic delay... when Intel deployed fixes for Spectre and Meltdown it crippled the performance of many Intel CPU's. Possible the same thing could happen to Intel in a couple of weeks? This way the new AMD CPU's would look even more favourable when launched later, instead of now. Anyone looking to upgrade to 9000 in the future would likely be looking at launch reviews, kind of makes sense.
I honestly doubt there's even a real issue at all. They are just doing 2 things. First they are buying time to have proper stock, and second, they are trying to look like they care about their customers just to pull as much intel users as possible.
@@WrexBF If the issue is as they say it is, of course two weeks makes a difference. If they found that some samples didn't perform to spec, but most do, they can revalidate and send out those that are fine - or just send out the next batch and work on revalidating the first ones later.
Lol well the 9900x literally performs identically to the 14900k in gaming but no, that couldn't possibly be why they pulled the chips already. Funny how a tech tuber in Italy posted a 9900x review and showed what I said earlier and they pull the chips shprtly.after that video released.
It's evident both intel and AMD seem to let the marketing/sales dictate the pace, at a cost to their QA. I refuse to believe smart engineers in their respective companies would overlook such problems unless they were overworked, or their concerns were falling to deaf ears
This is good news for pre-built PC companies! Many of our first PCs were pre-built, often during an awkward teenage console phase, and I'd be pretty upset if my first PC had a busted Ryzen or Intel CPU in it.
Seems like us consumers are better off waiting 1 year from the launch of any brand cpu to see how the real performance and stability really is. I think this is a new rule I'm going to adopt as neither company can be trusted at this point.
Definitely NOT the time to be fumbling with a dodgy launch. Good choice on AMD's part, even if annoying (assuming this is a good faith effort, of course. I don't really think they'd piss off their partners with this internal recall just to get a review edge over Intel)
We may be getting closer to the limits of silicon than we think. Mobile is doing ok but desktops seem to be running into problems with extremely high power and small process size
No, the problem is both AMD and Intel are just targeting moar clocks and no massive Zen 3/Alder Lake IPC improvements. The M1 Max made the 7700X look like a Bulldozer chip, both use TSMC 5nm and have similar perfomance but one has twice the power draw.
@@saricubra2867ARM and x64 are completely different coding types and architecture which is why you need emulation which makes it slower for the vast majority of programs
@@saricubra2867 One is also significantly bigger and significantly more expensive than the other. It's easy to make a fantastic CPU on the latest node if price and yields are of little concern.
@@saricubra2867 The Ryzen 7700X, is a medium range CPU, with 35MB of cache and 8 Cores, the M1 Max, is a high end CPU with 96MB of cache and 20 cores (divided like Intel as P cores and E cores) so both CPU's are a priore incomparable on power and market target alone! But there's other reasons, why no one should compare Intel and AMD CPU's, to those of Apple: One the M1 is a ARM based CPU, Intel and AMD are X86-64 CPU's, both are very different, bigger and more complex than ARM CPU's and not really 1 to 1 comparable, there's a reason ARM CPU's, were historically only used on cellphones and mobile machines and X86 only on desktop machines, because ARM was smaller, less complex and so more power efficient. Thou it has evolved since to became a viable desktop CPU as well. Two Intel and AMD CPU's, are made to work on millions of generic and different hardware configurations and memory speeds, they cannot optimize things very well, while Apple M1 CPU, is made to work on a single type of hardware configuration, where every component is chosen by finger and tested by Apple to work well together, is the diference between a game Console and a PC, a game console can do almost what the PC does but with less hardware, this is because it has static hardware and the software is easily optimized to run on it. And so, IMO is unfair to claim Apple CPU, is superior to Intel's and AMD's, not only because of what I said before but also because, it has no X86-64 CPU to really compare against the others. If Apple M1 was really such a impressive improvement over Intel and AMD offerings, enterprise CPU buyers would jump ship in a jiffy, but they are not, they know this stuff is not comparable. Enough said.
Better to delay than, as you mentioned, ship units with QC issues. I'd rather have something that works for the foreseeable future. Hopefully with the delay, you guys have your hands on production-ready chips in a reasonable amount of time to allow for proper testing. Kudos for asking them on the golden sample media lottery haha. Fingers crossed that Intel pushes the micro-code update prior to the 9000 series testing so we can get a true apples to apples comparison in the first take. Thanks for the update!
A delay of two weeks won't fix anything. It's like delaying a AAA videogame by two weeks. Also, Intel said ''sometime in August''. AMD gave the exact release dates (August 8th and 15th).
Was about to build 79something rig so I can retire this one to behind the TV, Zen 5 released announced. Stopped plans. Zen 5 release delayed, stopped plans again. Just gonna paint my flat I think, it could do with some redecorating...
Give AMD some props, when they found a problem rather than shipping it and letting customers deal with the fallout, they just pulled the whole smash and are running tests to make sure that unlike Intel you get what is advertised without the CPU turning into wasted sand after a few months. Kudos AMD, that is how a tech company should handle such a problem 👏👏
@@erkinalp Negative. A CPU that comes out of a prolonged idle and executes a math-light single threaded application HAS to hit max boost for at least a short blip. Otherwise "up to" is false marketing.
@@andersjjensen Agreed, if you advertise a boost speed, it must be reached. At least under certain circumstances. Not sure if you advertise a single-core boost of let's say 5.5GHz and dual-core boost of 5.4, but the CPU more or less always has two cores under load (and Windows always puts load on cores) how that would be seen.
The world isn't going to fall apart if Zen 5 CPUs aren't available for an extra few weeks. I'd rather products get delayed then have them early and have issues.
I guess they could sell them as less expensive SKUs, e.g. non-X I don't understand how they left the factory if they weren't up to spec, though, aren't they validating them before shipping?
What it could do is, suppose the units they have shipped end up with, say, 10% not even getting to the base clock. This delay will allow them to re-evaluate all units to catch all of those that are unable to meet their rated specs. They can do extra quality checking and replace the bad ones that slipped out with newer units that are in spec. Of course that 10% was a completely made up number, and it is possible that this isn't a failing to meet base clocks issue. There are plenty of other issues, like not much or no boosting headroom, signs of exploding processors or more.
So, what AMD is saying: We clocked our CPUs too far out of the box, which lead to them being unstable. This is the only case I can think of that could be fixed in just 15 days.
Or, "oh sh*t, I wonder if we have oxidation probs too". Who knows? July 31st already hinted at some sort of delay to me. Who launches a new generation of product on the last day of the announced month?
Or "We binned them a little too loosely, so now only 99.5% will actually hit target with the latest microcode, and we'll rather recall now than have to revert the optimizations". There is a lot more to CPUs than just frequency.
@@systemBuilder it's likely binning . 9900x/9950x are higher bins, and they probably didn't QC them quite right, so sent the other ones back (probably a few failed stock settings)
Smells of anti-competitive backroom deals with Intel. It was already suspicious in AMD/NVidia side, but looks like there's AMD/Intel backroom deals, too.
Is the general consensus that a delay seems better than a botched launch? I'm sure AMD was looking at Intel's situation and considering the optics of their own problems.
Help support our independent reporting! Grab a PC building anti-static Modmat here: store.gamersnexus.net/products/large-modmat-gn15-anniversary
Or one of our project & soldering mats! store.gamersnexus.net/products/gn-project-soldering-mat
Learn more about the Intel situation here: ua-cam.com/video/OVdmK1UGzGs/v-deo.html
Read our ad-free article! gamersnexus.net/cpus-news/amd-delays-ryzen-9000-did-not-meet-quality-expectations (support via the store & Patreon!)
Yeah I'd prefer a delayed Launch over dying CPUs
Just like video games. We prefer a push back launch than a buggy mess day 1 where its unplayable.
@@Lynnfield3440 you will get both in one package with how the world is going nowadays lmao
Yeah, my immediate reaction was “at least they caught it before they shipped to customers”. I’m not too bothered as long it’s not customers stuck with the headache
Yes, Delay something until its done right. At the price of $600, people expect the product to meet expectations and work correctly, if it needs another month, or even longer. They should take it. Much better than selling chips that can't handle such expectations.
A delay of two weeks is infinitely more preferable to nebulous issues down the line. No one will remember this in a few months. Commence unclutching pearls.
EDIT: to be clear, I was not suggesting Gamers Nexus were the ones clutching pearls.
Definitely agreed on the point that a proper launch will see these things forgotten, just like game delays.
But what can they possibly fix in such a short amount of time thought? Can't be a hardware issue, possibly just a software/bio fix or something, interesting nonetheless.
Something like this won't be remembered no but an actual issue will be remembered for years yeah
@@GamersNexus star citizen says hello lololol
@@SteveBennet500 Maybe they already fixed it, and tried to stay on schedule. Now they found out they can't, hence the delay.
Makes sense for AMD to take their time with this, especially since the competition isn't so compelling right now.
Definitely seems better to hold and get it right.
This is almost too good for AMD. By holding off and doubling down on production performance perfection as a poke at Intel, is straight up publicity gold for them!
@@GamersNexus Are the CPUs not already in stores though? So holding off doesn't really make much difference, or am I missing something?
@@SimBunkerthat was addressed in the first minute, retailers are getting replacement units
@@SimBunkerThey aren't. If they were, that would mean they were already launched (when launch was supposed to be end of this month) and then they couldn't be delayed, could they?
If by saying "in stores" you actually mean "in warehouses": That still means they aren't in stores and can thus be delayed.
What was your train of thought there, mate?
Intel: delays performance-botching patch until after AMD's launch.
AMD: delays launch until after Intel's performance-botching patch.
And who said performance will be worse?
@@Behemoth33lmao
you can't make this stuff up lol! it's hilarious to see
@@Behemoth33 everybody? That's how these things always go, security patch, tank performance. Stability patch? tank performance.
The reason it's unstable and degrades is that it's pushing too much for performance
@@Behemoth33 because like always intel knows only 2 ways to fix problems, increase wattage or decrease wattage
“A delayed CPU is eventually good, but a rushed CPU is bad forever”
~ Shigeru Miyamoto, 1913 - 1914
The wisest 1-year-old I ever met. 😔
@@benjaminoechsli1941I don’t think the age is the issue…
No Man's Sky has entered the chat.
"Don't believe everything you read on the Internet." - Genghis Khan
"I totally said this"
-Shigeru Miyamoto, 8000 BC - Present
Absolutely fine with a delay over quality issues. That actually makes me more excited for the new chips, since AMD at least appears to be focused on delivering a quality product.
What if all of this is just a marketing stint? There's nothing wrong with Ryzen 9000, they just wanna throwing shade on Intel.
@@fajaradi1223 I think it's most likely they want media outlets to test the new Ryzen against Intel **after** the patch, which in all likelyhood will decrease performance. This is Intel trying to squeeze by and AMD saying "denied."
@@fajaradi1223 Then good on them, for seizing the opportunity.
Competition like this ain't friendly, they'll push this as far as it can go.
How much can realistically be done in two weeks?
@@chris61986with coffee, Red Bull, and various stimulants it’s a lot of time lol.
AMD: Delays launch to meet quality expectations
Intel: Delays acknowledgement of issues to meet sales expectations
''Delays launch to meet quality expectations'' That's exactly what videogame companies say when they delay a game by two weeks. Those two weeks never change anything.
Difference between being a Billion Dollar company and a multi Trillion Dollar company.
This would not be funny if it were not plausible.
@WrexBF Of course it makes a difference.
That difference being a pile of files that insta BSOD's, and something that resembles a somewhat functioning piece of software that may be a game someday lol.
@@WrexBF how do you know this? Have you gone round and played a good number of games at their planned release date and then their actual release date and compared them?
I love the fact that Steve simply asks the question, "trying to find gold samples for reviews?"
They answered honestly too. AMD calls them platinum samples.
I like it also, but it's not like amd nor intel would answer truthfully to this question anyway.
It's in their best interest to not ship golden samples cuz it'll quickly be found out. But it has happened in the past, so it's a valid question. But i doubt either side would answer truthfully on that question.
But ye since they're recalling a whole lot more than review samples, i guess i could believe them somewhat on this one.
I’m confused, why didn’t they just increase the voltage of early chips to degrading levels so they could hit performance/stability targets? That’s what Intel does and that’s been going well for them!
because AMD is led by a real CPU engineer
Because if AMD did it all the Intel fan boys would wet there pants 😂 also AMD knows what happens if you fluff a cpu (Bulldozer and piledriver anyone) nearly went bust over it and lost a lot of good will, am 99% sure intel legal bills will put them at rist of going under as its going to be billions in liabilities esp for servers and OEMs Dell and HP ech are not going to want to recover there costs for sending engines out to companys to replace cpus as well as reputation damages
Posts at 1.6GHz, ship it already!
@@erkinalp That burnt holes in customers 7800x3d.
@@kramnull8962this one was actually mobo fault, and even then they just acknowledged it, refuded or replaced the chips and done. Sometimes things go wrong and it happens, it's how companies react to it that truly matters.
As someone that had to do a RMA of a 1800x due a memory bug that affected some of the inicial production units, finding this before retail sells them is a good thing.
Zen 1 was such a rough time not just because of all the issues but because absolutely nobody had any experience with the platform so most advice you received for issues was like alternate medicine tier placebo effect stuff lmao.
Also good idea to never buy at release, give it few months and you are more likely to get 2nd revision
@@Navhkrin was gonna buy a computer either way, I don't regret a thing. I am now on a 5800X3D which I'll probably keep for a few years.
@@rain8478 Zen 1 was rough in general for its whole lifespan. Mfgs didn't take Ryzen really seriously until the 3000 series. The motherboards were lacking so many features compared to Intel that you really gave up a lot of stuff for those extra cores and threads.
@@rain8478 i’m at the same boat. 1800x -> 3900x -> 5800x3d, happy with the decision at the time and don’t feel the need for a plataform upgrade any time soon.
"out of an abundance of caution, we're going to delay until Intel makes all of its CPUs slower"
it is awfully convenient for sure
AMD 4D chess
now, now, single thread turbo speed is all intel has
Found the Intel shill…
AMD: we'll wait for the intel patch for the reviews 🧐
Probably this
good point, with the patch Intel will be far behind
Certainly a possibility
Adjusting pricing....
Exactly what I thought
You know what? On the one hand I'm super disappointed because I'm overdue a CPU upgrade, but on the other hand, I would rather that companies release products when they're ready as opposed to when they look best on a balance sheet.
Same, but I’m getting the X3D. 9000 at best will be just as good for lower price
Even software companies that used to "release when it's ready" are under pressure and time crunch to release on a schedule. Probably hoping to fix on a XX.1 or z-stream release.
Are you really? What's on your PC at the moment?
@@junko4166 8600k lol
@@junko4166 I'm currently rocking a 5900X with 32GB of RAM that refuses to clock above 2433 (in spite of the advertised 3800mhz), and I'm doing a lot of AVX workloads where Ryzen 9000 should be a monster comparatively, and I want to upgrade to 192GB.
It's not really for gaming so I actually do want to upgrade very soon, and I honestly need the RAM, too.
AMD learning from others' mistakes. :)
Couldn’t do it any better tbh :D
Yep that's why it has all the issues lol
Yeaaahh, and also waiting to see how far they can jack the price up while the competition has an increasingly bad year...
Now if only their GPU division would do the same
They are just waiting to see how high they can price it.
Imagining the logistics involved in pulling products back, evalute and ship out again, 15 days sounds like an impressive turnaround time. AMD must be super confident it's just the first batch.
Careful there. That extrapolation may feel good, but it could just as easily be untrue.
They don't evaluate and ship them out again in those 15 days, they ship out completeley new CPUs that were already evaluated correctly. The old CPUs then can be re-evaluated at any given time
@@TheTaxxor True, obviously they could supply replacements even before receiving a shipment back. Yet, with linked reputational risks I'd expect they need to confirm via some re-evaluation before giving the go ahead. Probably they first retest trays, rather than ripping boxes open..
@TheTaxxor that would make sense, how do you ficks(tm) a busted finish product. Maybe rebadge them and turn off bad cores or something?
"Corrected an issue where we would have missed memeing on Intel's 'fixed' CPUs in benchmark graphs."
"A delayed CPU is eventually good, while a rushed one is forever bad."
-Shigeru Miyamoto, gaming industry legend
This speech is wrong in multiple levels
@@kirayoshikage9218 how?
i thought im finding Sun Tzu again
It is a famous Nintendo quote! Everyone knows it!
"A delayed game is bad, a rushed game is bad, all video games are bad, I fucking hate video games and making them."
- Shigeru Miyamoto, gaming industry legend
There were signs of it being delayed when they wouldn’t reveal the prices a week away from launch
True! Then they also made an "error" when disclosing the price of the 5900XT
They repriced RDNA2 halfway through review so not exactly new
@@spacechannelfiver the thing is: they had a price still. This time we just don't know.
They had to buy a 6800xt and trade their 6600xt to newegg. To do a little fine tuning.
Ha Ha the prices have just gone up as nobody will buy Intel right now! I hope Intel get through it as one CPU manufacture would be boring. Im in bed with both on my current systems but my gen 13 in mobile 😅
Well if there was a time for a delayed launch for QC reasons - this is it. Makes AMD look better.
You never know if they have long term quality issues until a year or 2 later.
you have to wonder if there even is a real qc issue for them to be seen to address
@@samsowden , look like. They pull all the products out of the stores, that costs money. This doesn't look like a fake delay.
Their chips blew up though, DDR5 platform has been a failure outside of 12th Gen
Just makes it look like they don't know what the hell they are doing. The reality is overall AMD has had far more problems with cpus and compatibility issues than Intel has had in the last 6-7 years.
*Pedantic post* of the day: August *16* is as mid-August as it can get. There are 15 days before the 16th in August, and 15 days after.
August 15.5 is half 31.
They delayed it by 15 days? Oh the horror.
It is for reviewers hanging on those views
Fr, i still feel like it hasnt been that long since ryzen 5000 was released
We are so used to everything moving so fast in this industry, it starts to look really absurd when you step outside it for a while and come back
@@pituguli5816 Same views either way, just shifted a couple weeks. No loss.
@@pituguli5816 No one is going to have the CPUs regardless, so what does it change for reviewers exactly? Plus, I don't think AMD really cares if a few dozen reviewers are getting inconvenienced compared to the millions of customers.
Very sus imo
LMAO, the INSTANT Intel released their announcement, my first thought was "AMD should totally delay their release just to piss off Intel". Good to know I'm in tune with their management
Intel said ''sometime in August''. AMD gave the exact release dates (August 8th and 15th).
On one hand I would love for AMD to be shrewd enough to actually think like that, but on the other hand I'm also happy that they are not. They've always been "dorks" at marketing. However fun the thought is, I think there is a legit concern of something here. That we haven't seen pictures of Ryzen 9000 boxes at retailers yet, and the like, suggest that they have been on the fence about this one for a while.
@@WrexBF they could always say "wait, we need another week"
@@BoraHorzaGobuchul I don't redelay would be wise just to wait for intel. They are already going to beat their older gen
I'm sure Intel is relieved to know they might be able to release their patch before Zen 5 reviews
I do wonder if the Intel news did not happen...would this have happened ? :)
Also very curious about this! Another timeline knows.
@@GamersNexus The other timeline also has Larrabee.
@@TechyBen ouch 😅
@@TechyBen Also, in the other timeline Bulldozer was released on target (2 years earlier) and AMD crushed Intel years ahead of the curve.
@@Gabu_Bulldozer always was doomed to fail because of the slow floating-point speed.
I do validation for automotive chips and the timeline doesn’t surprise me too much. I’d speculate that there’s a process parameter that’s causing issues and wasn’t being reliably caught by the testers. So when validation does the last bulk unit screen on production parts, before launch, some units pop up bad and they have to raise a red flag. The failure rate also needs to be high enough that C-suite doesn’t think they can brush it under a rug. Chances are they updated the testers to catch the issue and affected parts will be binned down.
I read "testers" as if it were people doing the job, so the conclusion was a bit... surprising o_o.
@@TheLK641 whoops, yeah I meant the ATE systems. Figured most people wouldn’t know what those are but probably should have been a little clearer than “testers” lol.
Don't worry, it made sense once I reread it, I just find the idea of a company going around updating their employees' firmware very funny for some reason *-*
The ad for the AIO at the beginning of the video is so authentically honest and positive, with pertinent information about the quality of the main fundamentals of the product that one would be concerned with, and it may be the best UA-cam sponsor spot I've seen as far as making me want a product based on the presentation
i actually clicked on the link in the description just to see what the price was. $60. If I end up needing an AIO anytime soon I guess they'll be first in line.
I bought a Thermalright Peerless Assassin 120 SE about six months ago for around $40. It's been a fantastic performer for a very impressive price.
@@SLLabsKamilion i got the 36o rad from them when it first came out... after a youtuber said best new hone brand..its almost seems too good to true...the business model is they are taking noctua tech from their last gen and making it cheap.
Seems like GN gets to choose their own words, with the sponsor just taking a brief glimse over it and giving a thumbs up.
I am not interested, but it's good presentation.
Damn, they learnt from Intel's mistakes really quickly.
Hard to not. You have to be some sort of inverted genius to somehow not learn from that.
They were lucky. QA decisions are usually unpopular, but the timing of this almost makes it a marketing ploy. Almost.
@@edwardallenthree If it is a marketing ploy, it's a fair one. Even if they're just pulling the tail on Intel, for obvious reasons... technically all of these manufacturers should be taking this approach of double checking there are no issues before selling their product anyways.
@@ManuFortis Instead of having 7800x3d with burn holes in the center of the chip?
It's only Intel left now to learn from Intel's mistakes.
At this point the 2024 disappointment t-shirt is going to end up being a robe in order to fit the entire list on it….
😭😭😭😭😭🤣🤣🤣🤣🤣
Just get 4XL version so that you can fit it all!
😂
lmao true
I was thinking full wall tapestry, woven from the dreams of tech insiders.
I'd seriously consider buying a robe this year, please do this if possible, @GamersNexus
woah. Chip business is having a rough month. 😅
Like a flight delay. People rarely realize they might as well DIE if the flight is not delayed.
You will not die if your pc will glitch.
@@Lookingformorefun You might if it is the computer on your plane.
@@Lookingformorefunyou might if your computer is your brain rot editing and mass content farming device, keeping you off the streets in the absence of any marketable skill. otherwise; you’re correct.
9:36 Agreed. Especially with the community sensitized to issues, AMD needs a flawless launch to demonstrate superiority. Heck, this announcement is probably a fantastic PR move!
Oh shit, AMD preventing their "Intel moment" in real time?
yes, engineer led company doing engineer things
@@erkinalp Lisa Su 🐐
Took a year last time.
@@LaughingSkull451 100%!!!
@@erkinalp I am totally Team RED, and it's a decision I'm quite proud of. AMD is a great company, led by engineers, as you point out. That's a huge difference!
Delayed is always better than poor quality at launch. We are learning to prefer this in games software, so honestly glad to see AMD focusing on quality in hardware land.
Learning? We always prefered it that way, the internet and with it the ability to fix stuff later gave devs simply a tool to push stuff out before its cooked to perfection and maybe fix it after cashing in.
Sometimes delaying something removes the first-to-market advantage and it may not be worth issuing the product at all. However, I think AMD is smarter than that, especially immediately after Intel has ventilated its own foot.
@@Gnarfendorf true but our tendency to pre-order everything and threaten devs over delays certainly doesn't help. The latter may be a loud minority, but it's better to make it clear that we want quality over quantity or speed here.
AMD: "Hey guys, you sure our CPUs are not burning themselves to death? lets check again before launch ;)"
Proceeds to run MC server 24/7
Oh I bet it's nothing like that, since they were claiming that TDP was down. They may have the opposite problem where they set the voltages too low and the chips crash at different power profiles from lack of voltage.
Considering I have a 7950X3D that can't work above 4800mhz at the moment, I do wonder.
@@aegiltech What CPU cooler and motherboard are you running, I'm running the Ryzen 9 7950X3D, RTX 4090 FE on the AsRock X670E Taichi with an Arctic Liquid Freezer II 360 While gaming the 3DVCache cores consistently hit 5.25GHZ and the regular cores exceed 5.7GHZ, highest core temps I've seen is around 74C which is below the 89C TJunction Max at which point the CPU will start the throttle, it would appear your throttling due to cooling....
@@N7_Nate ASUS Extreme x670E with a ASUS 360mm Ryujinx. Was happily running at 6000mhz with 2x32GB GSkill AMPsticks (which are on QVL list) until recently. Have tried different sticks, no joy, have another system with almost identical specs running a X670E ProArt downstairs, its fine. System doesn't pass memory training at 6000mhz, or if it does, it bluescreens windows.
need to try swapping CPU's with other system but have been too lazy.
Almost all retailers and channel partners are learning about this from you :) Thanks for the transparency and communication, AMD!
in before it's pr stunt
100% of the time, with consumer grade products, a delay, no matter how long, is preferable to launching broken products.
Today, on how to shit-talk the competition without directly shit-talking the competition
Lmao AMD just being Courteous before they destroy Intel again.
@@bionicseaserpent So prepare for AMD's greed
in light of what happened with intel, this is a super responsible action from AMD.
@@damson3413 now if only intel could learn from intels mistakes 🫢
Yeah it sucks that they had the issue at all, but if anything it just makes me want AMD products more because it makes me more confident that they are going to catch and fix problems before it becomes my problem
AMD is stupid… just wait until warrantry is over and then say that user did use XMP to oC the memory so it is customers own fault!
@@lightward9487 So you're saying AMD should copy Intel's reason for denying RMA replacements
@@lightward9487well, if your CPU is running healthy you won’t really need to call for warranty
:)
$10 says that AMD is waiting for Intel to release the microcode update, so it can beat Intel harder than initially planned.
Intel has been winning for a while now. AMD needs this release to win and even if they do it’ll only be temporary as intels next chip set will again be on top
Edit : AMD Fanboys jumping off buildings and the r9 is trash
@@madclone84 ROFL, Intel hasn't been winning even with their biggest fanboys recently.
@@madclone84 Uhhh, is that you Userbenchmark?
@@madclone84 Right... LMAO
i say they waiting for intel to see if they can fix it and then price accordingly if intel cant fix, that means AMD has no competitor for a few months until arrowlake comes out. they will pull an nvidia and itll be super overpriced.
The Steel Man approach, You have to respect AMD for wanting to compete with Intel on their best footing and not their worst.
The downside of competition is that companies move fast and break things to stay ahead. The upside is that you get a more competitive product
Hats off and cheers to all the QA managers who make hard decisions every day.
QA managers: logs six dozen Jira tickets on Friday 6pm and leaves for the weekend.
Well, for a product better to do it a bit late but get it right, rather than just launch anyway and tank their reputation the way Intel has.
exactly this
literally cyberpunk moment
I mean admitting it upfront is a good idea instead of a year later.
Admitting part of it... They still haven't addressed 14th gen failures. Or why they sat on knowledge of this all this year so far, letting unfixed CPUs keep on oxidising
@@greebj true.
they should ship a defect sample and the consumer sample to reviewers to fully restore confidence
I remember when your channel had only 20k subs. you have come a long way gamer nexus
Computer hardware be difficult, Getting it right is the hardest thing in the industry. There is a MASSIVE reason why there really isnt more then just a few CPU brands.
yeah, copyright.
literally has to do with 86x licensing
@@inkredebilchina9699 not just IP, it'd be hard to build it right even if you attempted to copy an existing design
Just got a 5800x3d so I’m good. Still love the merch n videos Steve
5800X3D ,, OVER-rated WAY over-hyped , over-priced
@@tilapiadave3234 It's almost as fast as a 7600X in games, but can be run without buying a new motherboard or RAM, so a way cheaper upgrade for the millions that already own a AM4 computer. I don't regret buying mine one bit.
@@tilapiadave3234They sometimes go on sale and are worth every penny when they do, they are hyped because they are good. 😂
@@tilapiadave3234 same can be said about anything. That’s your opinion
@@tilapiadave3234
Since it's now an older chip, if you can find it for a reasonable price it's pretty great.
Especially if you're on AM4 and don't want to spend more to move to AM5.
August 15th rolls around and AMD could have double the inventory plus a huge killing over Intel if they cant fix there issue mid August. This is Gold
Not double but definitely closer to re-supply, so there will be less stock drop off. Win win because there will be a lot of new customers after Intel did this dumb stuff.
from grocery, vehicles , solar panels, now computers....we having shrinkflation now...everything lasting 2 years now
Hopefully these issues (on both sides) have some silver linings for consumers. It'd be nice to see the new AMD chipsets come out at the same time as these CPUs too
Intel currently going “they have validation checks??”, “they have quality expectations?!”
AMD is stupid… just wait until warrantry is over and then say that user did use XMP to oC the memory so it is customers own fault!
😂😂😂
@@haukionkannel So you're saying AMD should copy Intel's reason for denying RMA replacements.
@@Dozeball Some people are absolute cuckolds for Intel.
@@haukionkannelwhy would the CPUs die anyway
@@Dozeball
Indeed! It would make Sense!
😂😂😂
Actually a smart move by AMD. They want to make sure the new Ryzen 9000 CPU models don't end up with same horrible crash issues that plagued the Intel 13th and 14th generation CPU's.
A delay of two weeks won't fix anything. It's like delaying a AAA videogame by two weeks.
@@WrexBF You're not paying attention. AMD told that it's an issue with the first batch of CPUs that has already been resolved, so the delay is just to retrieve the already shipped units and replace it with fresh ones
You don't even have to go to Intel, AMD learns from its own mistakes. This could be similar to Zen 2 where the CPUs were not reaching the advertised frequencies by a little bit (like 50 MHz or so), that one was resolved by an AGESA update if my memory serves right.
@@WrexBFExcept CPU MICRO code is not a big game.....
@@TheHighborn it's more complex than most game from an engineering perspective
If AMD are telling the truth, that's a good call. They're probably stalling to have the Intel microcode update out until they let reviewers have the chips =D
Kind of a chad move, AMD seems to be learning.
LMAO AMD telling truth ,, are you living in a dru-g induced bubble?
No, there is an actual problem with at least some of the retail chips.
Its all about the sheets the fanboys care about most.
Now I really wonder how that one Italian reviewer got his hands on the 9900X and if that chip underperforming had to do with that.
Maybe they already knew but were thinking right till the end that things will change, and when they didn't, they had to delay ...
What's this? A company that knows that it needs a bit more time to resolve any potential issues with its product? Has the bar been set so low by others that this is not even considered the norm?
Yes
Burn holes in the release of the 7800x3d ring a bell.
@@kramnull8962 stop posting this under every single comment over here. You know very well that a) this is unrelated to this story, as it was a Mobo issues, the CPUs were fine and b) AMD did what Intel is incapable of doing: be open about the issue, fix it in a reasonable amount of time and make customers whole.
Boeing & intel execs in the club having a good old trash-talk session about AMD right now...
@@Hugh_IWhat you're replying to is something I see with fanboys of certain brands all the time, they reply to things constantly to spread half-truths relating to a competitor of a brand they worship.
I wonder if this is more to force Intels hand then anything wrong on AMD's end. Now Intel will be forced to push their cpu microcode patches first which will impact performance. AMD will have the luxury of benchmarking against the REAL performance of these Intel cpus and not some bogus pre-patch number.
This
There's no actual facts that the intel will perform worse. Remember, lower voltage = lower temps = longer/better boost
But I have benched my 14th gen, will see how it will compare with the microcode results, so now that I have the basic results I'll keep that rig off and wait for the update and retest it
@@Behemoth33 If it's the actual CPU core that's failing due to high voltage, Intel will likely have to pull their boost clocks down a bit to maintain stability at lower voltages. With how high they are pushing the clocks, I doubt they will be able to guarantee a 6GHz boost with even a very slight reduction in peak voltage. The default specs they launched the CPUs with were no accident, they knew they had to push the voltage to that extreme to hit their target clock speeds, it's not like they pushed the voltage just because they felt like it.
I don't think this hypothesis holds water, because from what the video is saying their partners are genuinely sending the CPUs back and getting CPUs in return, which is incredibly expensive. GN has got third parties on the record stating that they're sending them back, which you wouldn't need to do if it wasn't a real issue. If all you wanted to do was to delay, you could bypass that cost by saying the issue was something you could patch remotely.
@@neruneri We don't know how many actual CPUS are being exchanged although if the number is high enough I agree with you.
Glad I bought a 5800X3D on sale! Hoping to grind out as much as I can from my AM4 system.
No kidding. At this point I’m intending on keeping mine until I have no choice to upgrade.
Damn I work at a PC retailer and have had 9700X on hand since July 6th. Getting CPUs and GPUs a week or two before launch is normal but 3 weeks seemed early so I didn't expect this! Then again it did seem strange that 2 weeks later we still only had 9700X and no other SKUs so I guess that explains it.
And yet other retailers still haven't gotten squat...
@@andersjjensen I'm curious if we'll have to send them back
What a W for AMD. Even if it's primarily a PR move, it essentially draws the comparison between Intel that sold subpar/defective products to their users without saying anything, compared to a delay for quality.
Astute observation. There might not even BE an issue, it might just be PR as you said, but it sure does make AMD look good.
Probably not just PR if they're actually shipping cpus back to retest. It's not cheap. ( But probably cheaper than doing it after launch)
Keep in mind that this A) costs very real money, and B) massively inconveniences their partners who have to deal with sending back the CPUs and waiting for the replacements. I can see an argument for how point A is mitigated by the PR win, but that's really only true if you don't majorly piss off your partners in the process. You wouldn't want to involve all the systems integrators in a PR stunt like this, and it sounds like those companies are in fact involved and sending their CPUs back.
@@user-nl7hw7uh1m It wouldn't make them look good to be seen to be lying.
My guess would be that SIs started doing internal testing and didn't like what they were seeing and alerted AMD. AMD did the right thing here in any case, claw back and test was a good move.
They were hoping The youtube crunch on 13th and 14th would make some idiots trash their Intel processors and run to AMD while the prices are high. Remember 100% bad cpu every one is pushing....
Amd wants to sell their AM5 backstock now, not at black friday.
My release date Ryzen 5950X suffered from occasional WHEA error reboots (not even a BSOD) shortly after launch. Later BIOS updates mostly mitigated the issue and I could go months without an issue, but over the 3 years I used it it still occasionally had stability issues. I eventually got it replaced under warranty shortly before that expired. This didn't seem to be a widespread issue, and most of the chatter about it seemed to be shortly after launch so I suspected that it only affected very early CPUs.
Given the stability issues Intel is facing AMD may have caught a similar issue with Ryzen 9000 and decided that it was best to get out ahead of it.
How did you know it was a WHEA issue? Was your infinity fabric set past 1800? I have occasional instant shut offs with no errors shown in windows event viewer on same processor
@@vaels5682 Windows Event Viewer occasionally recorded a WHEA error other times it just rebooted with no error. I didn't have it overclocked besides XMP to run the RAM at 3200MT/s.
@@TCL987hmm, i'm running this sweet 3600mhz Cl14 kit (no tweaked timings just xmp). I'll have to find a good search query to cover potential WHEA errors see if i come up with anything.
My mobo has dual bios so maybe ill try a bios update ive been hesitant to get and can go back to other if needed (i read the most recent bios update had slight performance reductions for a security update i didnt see as being necessary)
@@TCL987thanks for info
I've read a fair share of user reports of WHEA errors on Ryzen 5000 CPUs, but luckily RMA was completely hassle-free in all cases, which is an aspect I need to compliment AMD for.
Sometimes I think the GN videos are too long where Steve just rambles on and repeats himself, like this one could be 3 minutes shorter and lose nothing. Just a thought. Cheers, keep up the good work!
Well, just read the press release then I guess lol
@marks9233 yes @gamersnexus is an evil mastermind !!! Homelander irl !!!!
Props to AMD for making a difficult decision for the greater good. Too many companies today would have just barrelled forward, hoped for the best, screwed over a lot of customers, then played the usual PR damage control card while continuing to collect sales revenue. AMD just delaying is the right way to handle a quality issue, and it makes me trust the company more.
What a fantastic time to have to upgrade my 5800x to meet the demands of content creation. Deal with rusty molten Intel's or settle for a "random Ryzen". Pain.
THE GREATEST TECHNICIAN THATS EVER LIVED
I hear you. Building a new content creation computer myself soon. The 7950x is looking pretty good today. But it's probably worth the couple weeks wait, if possible. Would be nice to see if the 14th gen Intel is truly fixed also.
'Random ryzen' lmao what? 7950x is going to be the easiest jump for you. If you game, get the x3d. I've built PCs for myself, FnF etc over 20 years, my 7800x3d build was the easiest one I've ever done. I didn't even have to install drivers ffs. It downloads them automatically now O_o. Rock solid from day one, quiet, cool, fast as. EXPO DDR6000 CL30 trains in under a minute on an ASRock x670e steel legend. AMD hit it out the park with the 7000 series and 9000 series looks to continue that. Zero issues, platform is stable as and I've been watching it closely since Zen 1.
@@N4CR "random Ryzen" was an attempt at a joke, eluding to the fact that quality could not be guaranteed, hence the delayed release. I don't want the easiest jump, I want the best long-term jump. The point of me waiting this long to upgrade was to not be forced to upgrade to a 2 year old Ryzen that performs worse than a 14900k in Adobe Premiere. It was to upgrade to a substantially better CPU like the 9000 series, or the 14th gen (if it didn't rust and cook itself). But now I have to wait until they figure out the 9000 series or fix the 14th gen before the 15th gen is released (assuming they didn't use the same flawed design in the 15th gen).
Imagine if it’s a PR stunt just to mess with Intel, and the processors are actually perfectly fine
I doubt it since AMD hadsimilar issues with the 7000 v-cache chips starting motherboard fires. AMD's issues were easy to fix, because TSMC makes good wafers, and it was all bios related. I think Intel has deeper problems and are pulling an ASUS move out of their asses.
Sure, cost them only about 12 billions of market cap, about half of their entire annual CPU revenue, so why not 😂.
exactly what I was thinking, it just seems too obvious for it not to be true
@@disco.volanteThat's basically nothing. It will easily rebound.
@@disco.volante market cap and revenue can never be used in a sentence like that 😂. Delete
I'm formally putting in a request for system idle power draw charts for the zen 5 tests, please :)
Could just be a strategic delay... when Intel deployed fixes for Spectre and Meltdown it crippled the performance of many Intel CPU's. Possible the same thing could happen to Intel in a couple of weeks? This way the new AMD CPU's would look even more favourable when launched later, instead of now.
Anyone looking to upgrade to 9000 in the future would likely be looking at launch reviews, kind of makes sense.
A two week delay can be forgotten. Broken CPUs will be remembered forever.
I'd rather get them late than get them not up to high standards
AMD's approach is literally the polar opposite of what Intel has done for their 13th/14th generation
A delay of two weeks won't fix anything. It's like delaying a AAA videogame by two weeks.
Probably delayed,to check up on their own microcode and to get mobo venders too play ball. They don’t want what happened to Intel happen to them.
I honestly doubt there's even a real issue at all. They are just doing 2 things. First they are buying time to have proper stock, and second, they are trying to look like they care about their customers just to pull as much intel users as possible.
@@WrexBF If the issue is as they say it is, of course two weeks makes a difference. If they found that some samples didn't perform to spec, but most do, they can revalidate and send out those that are fine - or just send out the next batch and work on revalidating the first ones later.
Lol well the 9900x literally performs identically to the 14900k in gaming but no, that couldn't possibly be why they pulled the chips already. Funny how a tech tuber in Italy posted a 9900x review and showed what I said earlier and they pull the chips shprtly.after that video released.
It's evident both intel and AMD seem to let the marketing/sales dictate the pace, at a cost to their QA. I refuse to believe smart engineers in their respective companies would overlook such problems unless they were overworked, or their concerns were falling to deaf ears
no? their marketing team has clearly all gotten lobotomies in the past year or so. Look at laptop CPU's. what the fuck is ryzen AI hx or whatever.
Yah, deaf ears for sure. I doubt the execs can hear anything that isn't AI right now.
also another benefit is that, they'll release closer to the launch of the new motherboards
as there was a rumor about them launching in like september
This is good news for pre-built PC companies! Many of our first PCs were pre-built, often during an awkward teenage console phase, and I'd be pretty upset if my first PC had a busted Ryzen or Intel CPU in it.
almost like shrinking the dies year after year comes with difficulties
Hey as long as AMD irons out any issues on release, I'm good with a delay.
Uh Oh!!! That's troubling!!! If there was a problem, good thing they caught it and HOPEFULLY, corrected it
It's PR, they are going "look at us, we care about the customer"
You are incredibly gullible and not even afraid to show it.
@@youdonegoofed LOL How is that?
@@bengrogan9710 no company care about customers neither AMD
@@computerscience1101 I never said they did, I said thats the image they want to project by doing this
Seems like us consumers are better off waiting 1 year from the launch of any brand cpu to see how the real performance and stability really is. I think this is a new rule I'm going to adopt as neither company can be trusted at this point.
Definitely NOT the time to be fumbling with a dodgy launch. Good choice on AMD's part, even if annoying (assuming this is a good faith effort, of course. I don't really think they'd piss off their partners with this internal recall just to get a review edge over Intel)
Thanks Steve! Happy to see the great coverage from you as always. Louis Rossman also shouted you out today.
August 16th is more the middle of August than August 15th.
We may be getting closer to the limits of silicon than we think. Mobile is doing ok but desktops seem to be running into problems with extremely high power and small process size
No, the problem is both AMD and Intel are just targeting moar clocks and no massive Zen 3/Alder Lake IPC improvements.
The M1 Max made the 7700X look like a Bulldozer chip, both use TSMC 5nm and have similar perfomance but one has twice the power draw.
@@saricubra2867ARM and x64 are completely different coding types and architecture which is why you need emulation which makes it slower for the vast majority of programs
We'll have to jump ship to RISC eventually. I'm excited for the day, although I know it'll kill the software world
@@saricubra2867 One is also significantly bigger and significantly more expensive than the other. It's easy to make a fantastic CPU on the latest node if price and yields are of little concern.
@@saricubra2867 The Ryzen 7700X, is a medium range CPU, with 35MB of cache and 8 Cores, the M1 Max, is a high end CPU with 96MB of cache and 20 cores (divided like Intel as P cores and E cores) so both CPU's are a priore incomparable on power and market target alone!
But there's other reasons, why no one should compare Intel and AMD CPU's, to those of Apple:
One the M1 is a ARM based CPU, Intel and AMD are X86-64 CPU's, both are very different, bigger and more complex than ARM CPU's and not really 1 to 1 comparable, there's a reason ARM CPU's, were historically only used on cellphones and mobile machines and X86 only on desktop machines, because ARM was smaller, less complex and so more power efficient. Thou it has evolved since to became a viable desktop CPU as well.
Two Intel and AMD CPU's, are made to work on millions of generic and different hardware configurations and memory speeds, they cannot optimize things very well, while Apple M1 CPU, is made to work on a single type of hardware configuration, where every component is chosen by finger and tested by Apple to work well together, is the diference between a game Console and a PC, a game console can do almost what the PC does but with less hardware, this is because it has static hardware and the software is easily optimized to run on it.
And so, IMO is unfair to claim Apple CPU, is superior to Intel's and AMD's, not only because of what I said before but also because, it has no X86-64 CPU to really compare against the others.
If Apple M1 was really such a impressive improvement over Intel and AMD offerings, enterprise CPU buyers would jump ship in a jiffy, but they are not, they know this stuff is not comparable.
Enough said.
AMD sitting there looking at what's happening with Intel right now saying "🤔....we got time."
I mean at least they're being upfront and honest instead of trying to hide it like intel
Some boxes on Ryzen 9000 had misspellings.
Wrong - unenforcible anyway - legalese somewhere in the packaging :p
@@Karibanu No, the boost number on the box said 5.7 when it is really 6.9ghz. They had to fix it!
Better to delay than, as you mentioned, ship units with QC issues. I'd rather have something that works for the foreseeable future. Hopefully with the delay, you guys have your hands on production-ready chips in a reasonable amount of time to allow for proper testing. Kudos for asking them on the golden sample media lottery haha. Fingers crossed that Intel pushes the micro-code update prior to the 9000 series testing so we can get a true apples to apples comparison in the first take. Thanks for the update!
A delay of two weeks won't fix anything. It's like delaying a AAA videogame by two weeks. Also, Intel said ''sometime in August''. AMD gave the exact release dates (August 8th and 15th).
Hurry up and launch so the 7950X prices can (hopefully) drop and I pick one up for my new build. 😉
It wont drop. Afaik 7950x already running out everywhere. Price actually went up in last couple months.
@@dudao4163
Just like with GPUs.
A 3090 has barely dropped in price, at least here in Europe.
Was about to build 79something rig so I can retire this one to behind the TV, Zen 5 released announced. Stopped plans. Zen 5 release delayed, stopped plans again. Just gonna paint my flat I think, it could do with some redecorating...
Give AMD some props, when they found a problem rather than shipping it and letting customers deal with the fallout, they just pulled the whole smash and are running tests to make sure that unlike Intel you get what is advertised without the CPU turning into wasted sand after a few months. Kudos AMD, that is how a tech company should handle such a problem 👏👏
Love how he had to rephrase the use of "mid". English language is evolving.
I appreciate that AMD is doing a pre-sale recall of potentially defective processors instead of inflicting users with defective units like Intel did.
Nah, AMD are waiting till Intel release their emergency, performance-reducing microcode in the next couple of weeks :)
Could be the whole point of the delay. Who knows?
@@ChrisM541 AMD did this because they ha actually had issues in them Lol
@@computerscience1101 "AMD did this because they ha actually had issues in them Lol"
--> ChatBullsh#tTranslate didn't work for ya, eh Ruskie?
As a normal consumer doing an RMA on a processor that is 75 mhz below spec is just crazy!
it's actually within spec as long as it can sustain base frequency at base voltage
boosts are goodies
@@erkinalp not if Intel advertize the boosts...then you paid for boosts too and are entiteled to have them as per specs
@@Ph3NiX80yeah, not sure what the other dude was on about
@@erkinalp Negative. A CPU that comes out of a prolonged idle and executes a math-light single threaded application HAS to hit max boost for at least a short blip. Otherwise "up to" is false marketing.
@@andersjjensen Agreed, if you advertise a boost speed, it must be reached. At least under certain circumstances.
Not sure if you advertise a single-core boost of let's say 5.5GHz and dual-core boost of 5.4, but the CPU more or less always has two cores under load (and Windows always puts load on cores) how that would be seen.
At least they caught the issue before release.
AMD: Oxidation DETECTED.
The world isn't going to fall apart if Zen 5 CPUs aren't available for an extra few weeks. I'd rather products get delayed then have them early and have issues.
This is interesting. Zen 5 was in production since about December iirc. These weeks won't magically create better silicon.
Seeing as how it takes about a month to produce one CPU the initial release won’t be new silicon.
I guess they could sell them as less expensive SKUs, e.g. non-X
I don't understand how they left the factory if they weren't up to spec, though, aren't they validating them before shipping?
What it could do is, suppose the units they have shipped end up with, say, 10% not even getting to the base clock. This delay will allow them to re-evaluate all units to catch all of those that are unable to meet their rated specs. They can do extra quality checking and replace the bad ones that slipped out with newer units that are in spec.
Of course that 10% was a completely made up number, and it is possible that this isn't a failing to meet base clocks issue. There are plenty of other issues, like not much or no boosting headroom, signs of exploding processors or more.
@@kotekzotThis is likely an issue that shows up after long term use.
Code mister Rizzler... Code
So, what AMD is saying: We clocked our CPUs too far out of the box, which lead to them being unstable.
This is the only case I can think of that could be fixed in just 15 days.
Or, "oh sh*t, I wonder if we have oxidation probs too". Who knows? July 31st already hinted at some sort of delay to me. Who launches a new generation of product on the last day of the announced month?
Nah AMD just wants to see better than intel by waiting for intel to release the new microcode which may lower intels performance
Or "We binned them a little too loosely, so now only 99.5% will actually hit target with the latest microcode, and we'll rather recall now than have to revert the optimizations". There is a lot more to CPUs than just frequency.
@@andersjjensen Also sounds likely.
Could be anything up to the packaging plant causing problems, doesn't have to be a design issue.
FYI here in the UK at least the 9600X and 9700X now 8th August while 9900X and 9950X on 15th.
This means there is something wrong with the switching fabric!
@@systemBuilderPossibly but the 9600 and 9700 are both low tdp. The issue is likely worse with the high power spec processors
@@systemBuilder No it doesn't. It makes it a possibility that it's specific to dual CCD layouts, but by no means certain.
boooom it's a binning issue. i suspected this in a few comments earlier lmao.
@@systemBuilder it's likely binning . 9900x/9950x are higher bins, and they probably didn't QC them quite right, so sent the other ones back (probably a few failed stock settings)
Smells of anti-competitive backroom deals with Intel. It was already suspicious in AMD/NVidia side, but looks like there's AMD/Intel backroom deals, too.
Let's hope the delay on AMD's side proves to be worth it.
Very interesting about the 9000 delay. Seriously hoping they haven't discovered via oxidation issues! Would be a lot longer than a 2 week delay if so.