Ryzen 7 9800X3D CPU Failure Cases Surpass 100 Instances

Vendor	Cases	Percentage
ASRock	98	82%
Asus	16	13%
MSI	5	4%
Gigabyte	1	1%

[-]

Just_Maintenance@reddit

I'm sorry but what's with AMD and Intel just having absolutely no control over their mobo partners?

[-]

ListenBeforeSpeaking@reddit

It’s a catch 22.

You either lock down the motherboards and make every board the same or you allow them to innovate and differentiate to compete.

If you make every board the same, eventually there will only be 1 or 2 mfg’s left.

[-]

burnish-flatland@reddit

That’s cool, they haven’t really innovated anyway. I’ll take a plain $100 mono from 2005 any time for the GaMinG junk sold today.

[-]

ListenBeforeSpeaking@reddit

I think you’d be surprised about the crazy things MB makers try to do to get that 1% benchmark improvement that shows in reviews that ultimately drives a ton of sales.

Everything from cherry picked clock generators to increased voltages to tweaked load lines.

They just started adding the stubless dimm slot contacts. Some have innovated with attachment mechanisms, massive traces, crazy VRMs, wacky heatsinks, PCIe bridges, etc etc.

The motherboard business is pretty brutal.

[-]

Vb_33@reddit

The power of capitalism in full display.

[-]

cadaada@reddit

What would be the alternative? One cpu maker, one mobo maker?

[-]

DinosBiggestFan@reddit

Your only option is green PCB with no toolless features or fun things like connectors on the backside. Your move.

[-]

azenpunk@reddit

My move is to call that a win.

No more manufactured trash would be nice.

[-]

You know you're in reddit when you make a comment about capitalism and it instantly gets assumed to be a negative criticism of it. No I'm saying competition is good, competition drives a healthy market and it's good to see companies doing everything they can to carve out a niche for themselves.

[-]

cadaada@reddit

No, you are in reddit where you say capitalism bad and you get thousands of upvotes, come on.

[-]

Strazdas1@reddit

the job of a motherboard is to provide the board for all components to work together. It should not have different performance. The competition should be in things like PCIE lanes, connectors, etc.

[-]

mundanehaiku@reddit

You can have competition in other forms of economies. Under capitalism we have excessive amounts of product segmentation, bullshit marketing wank, confusing specifications, and wasteful aesthetics.

It would be nice if there was a lot less marketing and incremental products. Society would be so much better if there was less time/money spent on getting people to buy stuff and if there was an easier way to understand what you're buying at what price. Instead to not get ripped off and get the features I want I have to consult some third party matrix while ignoring all the bullshit marketing buzzwords and while looking up local pricing.

[-]

DJKaotica@reddit

I think you’d be surprised about the crazy things MB makers try to do [...] that ultimately drives a ton of sales.

Just like the (sort of recent) news about ASUS basically denying many warranties these days? (Maybe it was just a few sporadic cases that were hyped up but iirc Gamers Nexus had a bunch of people post about ASUS denying their warranties)

I only bring this up because I used to pay the premium for ASUS mobos ever since I had an issue with my A8N-SLI Deluxe and had to RMA it after a couple months (still well within the warranty period) and they did it no questions asked.

Other than having to wait a week for a new board the RMA was easy and the experience was great. Admittedly they didn't really do much more than the bare minimum (honour their warranty within a reasonable time period when accounting for shipping), but they responded quickly and I didn't have to fight tooth and nail.

My 9800x3d build is the first build I've done since my A8N-SLI that won't be an ASUS board. That was late 2004/early 2005, so we're talking 20 years of ASUS only motherboards.

But the price premiums they are talking these days, alongside-them now denying warranties, just not worth it to me.

[-]

Jmich96@reddit

I can't imagine AMD telling motherboard manufacturers what voltages to run for stock speeds (default settings) would stifle innovation. Motherboard manufacturers would still be free to "innovate" increased voltages for non-default settings and to properly innovate in other areas.

[-]

Vb_33@reddit

To be fair it would stifle innovation of using default voltage as a differentiator. Specially when your board is more equipped to handle more than competitors. Not sure this matters to me but it does make sense.

[-]

Jmich96@reddit

it would stifle innovation of using default voltage as a differentiator.

I think the problem is that this shouldn't be the differentiating factor. The differentiating factors between motherboards should be featuresets, hardware, cost, warranty, settings that can be manually changed, and automatic tuning options (disabled by default).

[-]

Vb_33@reddit

Shouldn't be the differentiating factor according to who? You? They aren't your companies, you don't decide what should or shouldn't be. Again I don't care about this, it's not the reason I buy a motherboard. I'm just trying to be reasonable. In a free market if marketing more voltage gets you more sales then that's the right play whether we agree with it or not.

[-]

fuettli@reddit

In a free market if marketing more voltage gets you more sales then that's the right play whether we agree with it or not.

A "free market" where everyone gets to do whatever is not the perfect utopia you make it out to be. Being reasonable includes limits.

[-]

Vb_33@reddit

Thats beside the point.

[-]

fuettli@reddit

Beside what point?

[-]

Strazdas1@reddit

Shouldn't be the differentiating factor according to who?

basic common sense. Default voltages exist as they are for a reason.

[-]

Strazdas1@reddit

if not locking them down means CPUs fail or explodes, then please lock them down and let me buy a stable product.

[-]

Rare-Industry-504@reddit

You're really not supposed to "innovate" with voltage, my guy.

The partners are making Motherboards, not entirely new CPUs that can handle additional volts.

The CPU is what it is. You don't get to fuck around with it.

Motherboard vendors make accessories to go with the CPU, they don't make the CPU.

[-]

ListenBeforeSpeaking@reddit

They are “innovating” with Voltage in that they are trying to get an edge on competitors.

AMD gives MB makers a design guide and configuration manual for each CPU family.

If the spec is 1V +- 20mV, they attempt to be closer to 1.020V than 0.980V.

If the clock gen says 100mhz +-3mhz, they may push to 103mhz.

Asus has gone beyond that on occasion with certain parameters to try and get an edge.

The degree to which they provide individual voltage plane control on some models is also innovation. The “auto-overclock” routines fall into the innovation category.

There are a ton of ways motherboard makers can push the limits of specs, and still be within spec. They have on occasion exceeded specs.

[-]

Slyons89@reddit

I was under the impression that AMD tightened up the BIOS / voltage rules (or at least defaults) with the mobo manufacturers after the “exploding” 7800X3D issues right after that CPU launched.

So this is surprising and makes me think there is just a bad bug in Asrock voltage controller configuration or something like that causing massive spikes that maybe aren’t even being reported in monitoring software. Like it could be super-juicing the voltage during boot only or something like that.

[-]

Vb_33@reddit

If it's just ASRock then why is it happening to Asus, MSI and Gigabyte too? I'm sure there's some explanation why it's happening to ASRock more so say it's because they're the go to enthusiast AM5 brand due to their features like good bifurcation as well as price, but it's hard to say without more data.

One thing is for certain the common link here is AMD. If it was just an ASRock problem it shouldn't affect the other board makers, yet it does.

[-]

3G6A5W338E@reddit

due to their features

Such as ECC support.

[-]

troldrik@reddit

Yup, bought the B650E Taichi Lite BECAUSE of confirmed ECC support. Running DDR5-6000 with ECC using Asrock's Hynix timings with a 9800X3D.

[-]

quaaludeswhen@reddit

At 1.1V?

[-]

troldrik@reddit

Asrock settings has RAM at 1.4V.

[-]

Slyons89@reddit

There's definitely defective CPUs from every brand/type of CPU that has ever been created. Frankly ~100 reported out of however many tens of thousands (or hundreds of thousands) that have sold is not concerning.

But out of the reported cases, the vast majority of them being encountered on one brand of motherboard, with a minority from the other brands, is at least something to raise an eyebrow at.

[-]

RateMyKittyPants@reddit

Same thoughts. The number of cases sounds really low if you consider the count of chips and boards out there. Not that I have that number in my pocket but the limited supply of 9800s and ASRock boards for months on months is telling.

On the flip, Reddit doesn't account for all cases. Curious if a reddit pole for 9800 mobo brand would show a similar distribution to the failure rate. If 80% of 9800 owners are using ASRock, then that normalizes the data.

[-]

HotRoderX@reddit

Normally I would agree but the sample size to small and if we dismiss that.

The arugment that more people buy Asrock over the compaition is strong.

Asus has a horrible reputation online for CS

Gigabyte has a equally horrible reputation.

Asrock isn't really talked about much or least don't see them much.

[-]

Strazdas1@reddit

The arugment that more people buy Asrock over the compaition is strong.

really? asrock needs to be responsible for over 82% of all motherboard sales. I find that to be a very unconvincing argument.

[-]

rTpure@reddit

Do you have evidence that there are more asrock motherboards sold than all other manufacturers?

[-]

Slyons89@reddit

Hmm do you think Asrock is selling 98 times more motherboards than Gigabyte?

Or about 20 times more than MSI?

Or even 6 times more than Asus?

Highly doubt it. Asrock is a value brand and may sell a lot, but not that much more.

[-]

HotRoderX@reddit

True but what happens if we remove OEM boards from the equation.

I am sure any numbers posted include custom boards for the OEM's since there both major players and do make OEM products built to diffrent specifications then enthuist boards.

I am not trying to defend ASrock only looking at it from every angle instead of witch hunting which social media is known for.

[-]

PorTroyal_Smith@reddit

The arugment that more people buy Asrock over the compaition is strong.

https://www.reddit.com/r/ASRock/comments/1joah6o/ryzen_9950x3d_failure_with_asrock_x870_pro_rs_wifi/mkqda7t/

[-]

COMPUTER1313@reddit

And Asrock has been blaming AMD for the failed CPUs. That blame might have worked with Intel during the Raptor Lake fiasco (where CPUs were dying on all board brands).

[-]

Slyons89@reddit

Hmm I'm not sure Asrock has made any official statements about that, it seems more like conjecture from commentors on article and reddit up to this point. But I haven't deep dived into forum threads.

On one particular example they kind of finger pointed at debris in the CPU socket during installation as the cause of a failure, which is not blaming AMD directly, maybe user error or manufacturing defect on their part leaving debris in the socket.

[-]

Strazdas1@reddit

Board partners do things outside the specs and get away with it, so they grow bolder to do things even outside specs. Its actually not uncommon in the industry, its just that this is a very high profile occurence.

[-]

KennKennyKenKen@reddit

Intel and AMD gentle parenting ASRock

That's ok ASRock just do the best you can 🤱🤱🤱

[-]

ryanvsrobots@reddit

Funny how everyone is suddenly a statistician now compared to 12VHPWR failures.

[-]

Strazdas1@reddit

What was there to measure? 100% of 12VHPWR failures used the same connector.

[-]

ryanvsrobots@reddit

…the failure rate, very obviously.

[-]

Strazdas1@reddit

That we dont know because we dont know the return data. Based on data from one US shop that shared their data, the failure rate was about 0,05%.

[-]

ryanvsrobots@reddit

You’re the problem

[-]

user007at@reddit

Memory compatibility is definitely not the cause. Instead of investigating longer like Intel does, they push out a statement faster but it seems incorrect. I’d rather say they messed up something with power delivery.

[-]

ferpecto@reddit

As a new owner of 9800x3d Iam worried, using a Gigabyte mobo maybe slightly less so, but I don't want to double the number..

Still about a 100 cases out of 10s of thousands sold, and mainly it seems on ASRock, curious.

[-]

Slyons89@reddit

Run HWInfo64 and keep an eye on Vcore and VSOC.

Vcore (main CPU voltage for the cores) should ideally be staying below 1.4 V in most cases. On my MSI board, it rarely goes over 1.2 V (but I am running PBO Curve Optimizer with negative offset to reduce voltage). If approaching 1.5 V, that's worrying, but it shouldn't be.

VSOC (basically, power to the CPU memory controller) should be 1.3 V or lower in most cases (most boards will run 1.25 or 1.3 V by default for PBO but it varies). If approaching 1.4 V, that's worrying, but I doubt it.

Note: these 2 voltages are not the only things that can cause a CPU to fry but they are the 2 major ones.

[-]

ferpecto@reddit

Thanks, I tried it for a half day and I didn't notice it going up to those numbers so I think it should be ok. I uninstalled it cause for some reason I think it just randomly started causing my games to crash (after shutdown/starting again), maybe conflict with AMD adrenaline monitoring?? I hope so since it's been fine for hours now after uninstall.

I'll look into PBO curve optimiser.

[-]

iatelassie@reddit

Saving this for when I build my new computer this weekend (hopefully). Sprung for the fastest gaming cpu on the market for the first time in my life so I’m gonna shit myself and cry if it breaks.

[-]

zainfear@reddit

Yep and remember you can manually set these voltages in the bios. Thus relieving yourself from worrying about what the manufacturer's "auto" setting considers safe.

I'm running ASUS B650E-E with 9800x3d and the following voltages:

VDDIO / MVDD / MVDDQ: 1.35V VSOC: 1.25V

From buildzoid's memory oc guide: https://www.patreon.com/posts/low-effort-rank-77403831

[-]

Death2RNGesus@reddit

The failure rate should not worry you, if you have any idea about failure rates about electronics you should be HAPPY instead, these CPU's have an incredibly low failure rate, way below the historical average CPU failure rate.

[-]

Techhead7890@reddit

Eh, when the failure ends up with burning and fire risk, I'm not sure I could say I'm happy.

[-]

Strazdas1@reddit

a CPU burning itself is not the same as CPU setting fire to your home. Unless you are using wooden flamable socket support or something.

[-]

imaginary_num6er@reddit

Thousands sold of what? 9800X3D processors or AsRock boards?

[-]

ykoech@reddit

Maybe many people have ASRock boards?

[-]

coolthesejets@reddit

86% of the failures are on asrock boards, you think asrock has 86% of the market share?

[-]

ykoech@reddit

Definitely not but worth asking. There could be an anomaly in the market with AM5.

[-]

MrDunkingDeutschman@reddit

Computerbase.de recently had a motherboard survey. Asrock had like a 15% marketshare among their DIY community.

[-]

I-never-joke@reddit

for 9800x3d builds?

[-]

DktheDarkKnight@reddit

If it was Asus or MSI you can argue that. They are usually the most popular brands when it comes to motherboards. But not Asrock.

[-]

chefchef97@reddit

You think Asrock aren't moving volumes in the thousands-tens of thousands?

[-]

DktheDarkKnight@reddit

Just that they are not as big as Asus or MSI.

[-]

croix_de_guerre@reddit

You misspellee overpriced

[-]

dfv157@reddit

Don’t let your biases cloud you. We’re on a PC specific subreddit, this community cares about price/performance and min maxing. But when you look at 4060s being the best selling GPU ever when it’s widely panned as the worst gpu release in generations, you start to realize regular people just don’t care and buy what they know. They know ASUS/GB/MSI, not Asrock.

For the retailer I know, Gigabyte and ASUS is outselling MSI, and they all have orders of magnitude higher figures than Asrock (AM5, grouped by brand)

[-]

EkeeB@reddit

You got a source for those claims on the motherboard market share? That might be true for gpus/peripherals, but for motherboards, I think it's different. Most people buy Asrock boards because they are cheaper than the overpriced competitor while also providing decent quality and no pcie lane sharing.

After initial launch, their am5 motherboards are constantly sold out while I've seen other overpriced competitor's boards sit on the shelf for months. The x870E Nova and Taichi seem especially popular.

But this doesn't excuse Asrock for putting the bare minimum effort in trying to solve these dying 9800X3D issues.

[-]

dfv157@reddit

no pcie lane sharing.

You better bet normies have no clue what this means and how it affects them.

[-]

nickkuk@reddit

Normies aren't buying top end parts and building them themselves though.

[-]

Neverending_Rain@reddit

After initial launch, their am5 motherboards are constantly sold out while I've seen other overpriced competitor's boards sit on the shelf for months.

That doesn't necessarily mean ASRock is outselling other motherboard manufacturers. I'm giving to make up some numbers for an example:

Say ASRock expected to sell 50,000 boards in a time period and Asus expected to sell 300,000 in the same time period. They'll both make roughly that many boards. But in reality 70,000 people wanted to purchase the ASRock boards and 280,000 wanted to buy Asus boards. It'll look like ASRock is outselling Asus because they're sold out while Asus still has boards on the shelves, but Asus is still outselling ASRock significantly with these hypothetical numbers.

You can't guess which company made and sold more motherboards just by looking at is and is not in stock.

[-]

burnish-flatland@reddit

Being sold out doesn’t tell much beyond not being able to meet the demand. For all we know, other manufacturers may have 10x more demand and produce 10x more boards to meet it.

[-]

Dester32@reddit

Except that asrock what consistently the recommended mobo manufacturer for the Ryzen 9000 series. Sure, thetr arr disproportionate issues with asrock mobos, but don’t act like they are not mainstream at this point.

[-]

dfv157@reddit

Yeah, recommended on places like this sub where people actually do their research and want the “best”. Unfortunately (fortunately in this case??) most people don’t need the best or bother doing research. Did you know B650s are the best selling boards still? What did you need a X870E for exactly? None of our recommendation is for a high end chipset on AMD unless they specifically mention in their use case to justify it.

[-]

HotRoderX@reddit

Asus and Msi have both hurt them selfs over the past few years in social media.. so it wouldn't be shocking for people to look for alternatives.

Asus with there bogus warranty issues that GN covered.

MSI with there scalping on the 4xxx gpu's.

[-]

Vb_33@reddit

ASRock is the go to AM5 board brand for enthusiasts, it was the brand always recommended at least until now. I know this doesn't equate to more sales but it's not like they're biostar or something.

[-]

scrndude@reddit

Asrock’s been the cheapest and most recommended boards for AM5, I think they might be outselling Asus and MSI.

[-]

dfv157@reddit

Not even close. They might be the best option for X870E on paper, but most buyers are just doing the $150-200 B650 boards (even cheaper with some retailers bundles) which are AM5 last time i checked.

[-]

FinalBase7@reddit

Pretty sure Asrock have been consistently making some of the best budget and some of the best high end boards while under cutting the competition, tho I agree, they still shouldn't be selling more than Asus no matter how shit Asus mid range and budget boards are.

[-]

DktheDarkKnight@reddit

Exactly my point. I don't understand why people keep misunderstanding my statement. It's purely about sales. Am not saying one is better than the other.

[-]

uKnowIsOver@reddit

Misinfo, not AMD fault ofc.

[-]

imaginary_num6er@reddit

It’s obviously ASRock’s fault

[-]

nickkuk@reddit

Obviously not when Asus, Gigabyte and MSI had CPUs dying also. It's an AMD issue, I would put money on it being a v-cache flaw.

[-]

FinancialRip2008@reddit

i'm assuming it's an amd issue that asrock's design philosophy is highlighting. asrock tends to fly closer to the sun when it comes to cheaping out on components that shouldn't matter assuming everyone else is to spec.

[-]

demonstar55@reddit

What's the batch number? I haven't updated my BIOS because of this, it's still working, but I want to update my BIOS, especially since Gigabyte for some unknown reason removed my version for download (also same mobo as the 1 Gigabyte example)

[-]

iatelassie@reddit

You didn’t update bios for the 9800x3d? I’m getting a gigabyte board this week so am just curious. And they listed the batch numbers of the cpus…seems to be from two batches.

[-]

demonstar55@reddit

I updated the BIOS when I first set up the system. No idea what BIOS it came with, but I updated to a BIOS that most certainly was released well after the board went to the store.

[-]

iatelassie@reddit

Ok yeah that’s what I’m going to do too. Was just checking - this is my first real build.

[-]

Numerlor@reddit

I struggle to believe this is entirely on the mobo vendors when there haven't been any failures like these before AM5. Their 1.3V "fix" was also obviously not an actual fix and just jumping on what was the popular consensus at the time or they wouldn't be still burning out

[-]

Renard4@reddit

If you look at the distribution across vendors then it should be clear there's something wrong with one of them.

[-]

Numerlor@reddit

Yes, but that doesn't necessarily mean the fault is compeltely on their side

[-]

NGGKroze@reddit (OP)

https://www.reddit.com/r/ASRock/comments/1i5iy9a/update_and_summary_on_the_dead_9800x3ds/ - the used source in the article is AsRock subreddit, started 2 months ago, updated right now, up to 108 cases

[-]

BookPlacementProblem@reddit

Well, that is indicative. AMD's business practice seems to be "Just a little less worse than our competitors."

[-]

ApplicationCalm649@reddit

Some day people will learn not to buy ASRock motherboards. Not today, but some day.

[-]

Kougar@reddit

As opposed to what, the early ASUS boards that slagged AM5 CPUs, then committed seppuku slagging the socket afterwards trying to power an already-dead processor? Or the early Gigabyte boards that also slagged AM5 CPUs, and had UEFI bugs (which got fixed, then unfixed in later UEFI versions) that voltage spiked still more CPUs? And those were across all chips installed in them, not a single specific model which is the case here.

[-]

Nvidiuh@reddit

Well, even though I haven't made a purchase yet, this makes me feel a bit better about having an MSI board on my parts list.

[-]

hurrdurrmeh@reddit

100 divided by however many sold seems like an excellent non-failure rate.

Well done AMD.

[-]

Sensitive_Ear_1984@reddit

100 reported on Reddit. Not total.

[-]

hurrdurrmeh@reddit

Just going off the title.....

[-]

SirActionhaHAA@reddit

On asrock boards. The others look normal.

[-]

COMPUTER1313@reddit

It’s still not a good look when only the CPU failures are on the Asrock boards, and Asrock has been blaming AMD for the dead CPUs.

[-]

nickkuk@reddit

That's not true though. The failures are happening on all manufacturers motherboards.

I'd put money on it being a v-cache issue, it is only happening with v-cache CPUs on a wide range of motherboards, X870E, X670E, B850 on every brand of mobo. Asrock are the most affected but not the only one.

[-]

jontseng@reddit

Quick thoughts.

So is 100+ failures a big deal? To determine that we need to figure out how many 9800X3Ds sold.

For context, AMD had $$2,313m of client revenue in the Q4 when the 9800X3D was launched. Desktop CPUs probably around 40% of that so you have something like maybe $900m of desktop CPU revenue.

How many of these are 9800X3D? It's a high end and therefore lower volume SKU. But of course its also very popular especially in the launch period (where you also have channel fill).

Finger in the air say a third of sales is this chip? I don't think its the majority of that $900m. But I don't think its like $50-100m either.

If we assume a $400 ASP (obviously retail is $500 but there is channel and disty mark-up) and $300m of sales you are getting to something like 750k units sold.

100 / 750,000 is like a 0.01% failure rate. Of course those 100 cases are not going to be all of the cases by a long-run, just the reported cases largely in English-speaking countries. But to put it another way - if the failure rate was to 1% then we would be seeing 7,500 cases.

This isn't to downplay the impact and there are a bunch of error bars around the estimate. But its helpful to try to set this into context.

Hope thats helpful!

[-]

TALMOR-187@reddit

100 cases posted on Reddit. That's it. God knows how many cases there are outside of Reddit.

[-]

jontseng@reddit

Yes that's why I back calced how many failures you would need for a 1% fail rate.

For that to be the case then you would need 100x the number of cases posted to Reddit (or to put it another way Reddit reports would need to be 1.3% of total reports).

This is of course based on my somewhat dubious unit extrapolation!

[-]

unknown_nut@reddit

At this point, not overvolting and not auto overclocking needs to be a bullet point for CPU buyers.

[-]

Silent_Fault2969@reddit

Quality of many type of products has been trending down since pandemic.

[-]

airfryerfuntime@reddit

How many of these have had the IHS pried off?

[-]

szczszqweqwe@reddit

Phew, good that I recently recommended Gigabyte's motherboard to a friend

Well, there were already Asrock/Asus rumors and at that price Gigabyte had better offer than MSI.

[-]

looncraz@reddit

That's really not very many considering the sales numbers. Likewise, I think ASRock outsold all other vendors for boards, so the ratio of failures might be similar across brands and it's just a normal CPU failure rate.

[-]

ConsistencyWelder@reddit

First off, didn't they say the CPU isn't damaged, it's a BIOS issue that can be fixed? But that the CPU's get misreported as broken.

And also, considering they've sold a billion billion of them, 100 of them failing wouldn't surprise me the least bit.

[-]

zainfear@reddit

No. In many of the reported cases there is actual physical damage visible. CPUs fried to death. There is no reason for that kind of damage unless the user has been overvolting the CPU to hell and back, which most aren't.

[-]

_OccamsChainsaw@reddit

I think there is a bias in that when something is reported on, it must mean it's an issue when literally everything has a failure rate, we're just not consciously aware of it.

9800x3d here with the same msi tomahawk board that the first reported failures were on. And a 5090 with all the ROPs and no melted cables. 🤷🏼‍♂️

[-]

zainfear@reddit

Sure, some people are just unlucky and receive monday morning samples.

But then you would expect the reported numbers of failures to correlate with mobo market share. I don't know Asrock's market share, but it is not 80%, like their share of reported cpu deaths.

[-]

EmilMR@reddit

regardless of whose fault this is, 3D cpus seems more fragile than others. Probably shouldn't expect them to last 10 years or whatever like back in the old days.

[-]

zainfear@reddit

CPU overclocking has always come with risks. The problem is these board partners auto-OC'ing with unsafe voltage limits to have a bigger bar in the benchmark graphs. Leading to customers wondering why their CPU exploded even though they didn't do anything.

[-]

NuclearReactions@reddit

Out of how many sold? I saw some figures from march talking about 9000 units but that seems incredibly low considering that everyone who is upgrading (that I know of) is going for the 9800x3d

[-]

TomSchofield@reddit

So, in Germany alone there were around 9000 sold in Feb, and that's only cos they're biggest online retailer. It's in 6 figures

[-]

ConsistencyWelder@reddit

IF you're talking about Mindfactory, you're right. They sold 6-9000 per week for a while, maybe still do. While they're selling a couple hundred Arrow Lakes in the same time.

[-]

TomSchofield@reddit

Yup, that's the one, couldn't remember the name

[-]

NuclearReactions@reddit

That makes waay more sense

[-]

FinalBase7@reddit

I don't think CPUs tend to fail after initially working tho, Dead on arrivals are not counted in this, I would even say a 0.1% failure rate on CPUs that worked on arrival is too high.

[-]

Yourdataisunclean@reddit

Does anyone have any numbers of how many of each mobo/CPU is sold? I see people panicking/dissing asrock. But without the overall numbers to put things in context its hard to determine if a 0.1%, 0.01%, etc failure rates is really worth getting concerned over.

[-]

6198573@reddit

I don't think anyone has any good sales numbers

We can only draw conclusions from:

the rate at which these failures are getting posted on reddit compared to other CPU launches
the difference between the number of asrock posts vs other brands (so far they account for 80% of the failures but there's no way they sold 80% more than other board partners)

[-]