News Intel 13th Gen CPUs allegedly have 4X higher return rate than the prior gen — retailer stats also claim Intel CPU RMAs are higher than AMD

[Admin]

A European retailer reported that Intel processor RMAs have jumped from 1.75% to as much as 7% in recent years.

Intel 13th Gen CPUs allegedly have 4X higher return rate than the prior gen — retailer stats also claim Intel CPU RMAs are higher than AMD : Read more

 

[WDPowell]

"Intel has recently been making rounds in the news with instability with Raptor Lake and Raptor Lake Refresh processors. While the company has announced a mid-August patch to address the crashing and instability problems, the software update won’t repair any chip that has already died or is dying. For example, one of our Intel Core i9-13900K chips refuses to work with Nvidia’s graphics drivers and crashes when some games are launched, and we had to RMA that chip."

I have an Alienware M18R1 that had to get service for a similar issue. Any attempt to launch WoW would initially cause a memory error. Enough attempts would ultimately result in a blue screen followed by a reboot. Once the machine came back up, nothing worked until I manually shut down or rebooted the laptop. For example, browsers would load blank pages and do nothing else. During the troubleshooting process, I tried updating my Nvidia drivers only for the installer to fail to detect the 4090. Once I rebooted the laptop, the installer was able to detect the 4090. I originally thought it was a RAM issue or a faulty motherboard, since the motherboard had been replaced during a previous RMA request. Now I'm suspecting the CPU was failing due to what is being widely reported with the i9 chips.

My concern now is whether my current board/cpu has sustained damage. Fortunately I have a several year warranty on it.

 

[TheHerald]

"Intel has recently been making rounds in the news with instability with Raptor Lake and Raptor Lake Refresh processors. While the company has announced a mid-August patch to address the crashing and instability problems, the software update won’t repair any chip that has already died or is dying. For example, one of our Intel Core i9-13900K chips refuses to work with Nvidia’s graphics drivers and crashes when some games are launched, and we had to RMA that chip."

I have an Alienware M18R1 that had to get service for a similar issue. Any attempt to launch WoW would initially cause a memory error. Enough attempts would ultimately result in a blue screen followed by a reboot. Once the machine came back up, nothing worked until I manually shut down or rebooted the laptop. For example, browsers would load blank pages and do nothing else. During the troubleshooting process, I tried updating my Nvidia drivers only for the installer to fail to detect the 4090. Once I rebooted the laptop, the installer was able to detect the 4090. I originally thought it was a RAM issue or a faulty motherboard, since the motherboard had been replaced during a previous RMA request. Now I'm suspecting the CPU was failing due to what is being widely reported with the i9 chips.

Nah, that isn't a CPU issue.

 

[Taslios]

"Intel has recently been making rounds in the news with instability with Raptor Lake and Raptor Lake Refresh processors. While the company has announced a mid-August patch to address the crashing and instability problems, the software update won’t repair any chip that has already died or is dying. For example, one of our Intel Core i9-13900K chips refuses to work with Nvidia’s graphics drivers and crashes when some games are launched, and we had to RMA that chip."

I have an Alienware M18R1 that had to get service for a similar issue. Any attempt to launch WoW would initially cause a memory error. Enough attempts would ultimately result in a blue screen followed by a reboot. Once the machine came back up, nothing worked until I manually shut down or rebooted the laptop. For example, browsers would load blank pages and do nothing else. During the troubleshooting process, I tried updating my Nvidia drivers only for the installer to fail to detect the 4090. Once I rebooted the laptop, the installer was able to detect the 4090. I originally thought it was a RAM issue or a faulty motherboard, since the motherboard had been replaced during a previous RMA request. Now I'm suspecting the CPU was failing due to what is being widely reported with the i9 chips.

My concern now is whether my current board/cpu has sustained damage. Fortunately I have a several year warranty on it.

... It may not be cpu.. but i'd still open an RMA for it. Your errors sound very similar to the sorts of things that have been reported as being part of the instability.

 

[JarredWaltonGPU]

"Intel has recently been making rounds in the news with instability with Raptor Lake and Raptor Lake Refresh processors. While the company has announced a mid-August patch to address the crashing and instability problems, the software update won’t repair any chip that has already died or is dying. For example, one of our Intel Core i9-13900K chips refuses to work with Nvidia’s graphics drivers and crashes when some games are launched, and we had to RMA that chip."

I have an Alienware M18R1 that had to get service for a similar issue. Any attempt to launch WoW would initially cause a memory error. Enough attempts would ultimately result in a blue screen followed by a reboot. Once the machine came back up, nothing worked until I manually shut down or rebooted the laptop. For example, browsers would load blank pages and do nothing else. During the troubleshooting process, I tried updating my Nvidia drivers only for the installer to fail to detect the 4090. Once I rebooted the laptop, the installer was able to detect the 4090. I originally thought it was a RAM issue or a faulty motherboard, since the motherboard had been replaced during a previous RMA request. Now I'm suspecting the CPU was failing due to what is being widely reported with the i9 chips.

My concern now is whether my current board/cpu has sustained damage. Fortunately I have a several year warranty on it.

FYI, in my case, there were initially instability issues with shader compilation in certain games, as well as in CPU stress tests like Cinebench multi-threaded. I was able to work around those to some extent, and eventually found BIOS settings that appeared to work. That was in ~February. I started retesting all GPUs. But by May I found that I couldn't update my Nvidia drivers — the GPU was detected fine, but the installer would just fail.

At this point, I tried a different motherboard, RAM, updated BIOS settings, and even did a fresh install of Windows 11. That last actually crashed a couple of times after the initial file copy, meaning during the "detecting devices, etc" stage of the install it crashed. It was able to recover and eventually get Windows 11 up and running, but I couldn't install Nvidia drivers and some other apps were also unstable. That's when I RMA'ed the CPU.

 

[WDPowell]

... It may not be cpu.. but i'd still open an RMA for it. Your errors sound very similar to the sorts of things that have been reported as being part of the instability.

It's been RMAed and the board had been replaced. This particular board was itself an RMA replacement, which I suspect was a refurbished board. We were able to rule out RAM since Dell did send a technician out to replace my RAM. Unfortunately the issues persisted. It could have been a GPU issue, but I found it interesting that my CPU was an i9 and some of the symptoms matched what the article mentioned. This issue was something that seemed to develop over time. When I got the laptop back, from the first RMA, I was able to game on it just fine. I took an extended break from said game then couldn't launch it anymore when I tried picking it back up.

 

[magbarn]

Intel needs to do a FDIV type of recall on these chips or at the very least double the warranty to 6 years to rebuild confidence. I've lowered the power limit and voltages on my 13900K and seems to be stable now, but I'm not doing RMA as long as it's stable and the "fix" hasn't been fully tested in the wild.

 

[Charizor234]

Intel needs to do a FDIV type of recall on these chips or at the very least double the warranty to 6 years to rebuild confidence. I've lowered the power limit and voltages on my 13900K and seems to be stable now, but I'm not doing RMA as long as it's stable and the "fix" hasn't been fully tested in the wild.

I'm in the process of RMA-ing my 14700KF right now. I'm just hoping it didn't mess up my board because it's been faulty and running way too hot for over 6 months now. It won't function at the recommended voltage but works if it's run at a higher voltage even then though, it overheats instantly.

 

[Zerk2012]

Part of the article just doesn't look right.

"with some institutions reportedly jumping to AMD’s Ryzen 9000 CPUs.

AMD has been having some issues on its side as well, with the launch of the Ryzen 9000 chips pushed back a couple of weeks due to an unspecified issue"

How can you be jumping to something that's not released yet?
Why would you be planing now to buy something that is already having problems without knowing it will be fixed properly?

 

[Geekaycee]

This is just a start.

 

[greymaterial]

I have an Alienware M18R1 that had to get service for a similar issue.

is your cpu i9-13900HX? this cpu has the same B0 die as 13900K/KS/KF, the same desktop chip used in mobile configuration. did you use AWCC "performance" mode or OC? those would accelerate the issue as well.

 

[WDPowell]

is your cpu i9-13900HX? this cpu has the same B0 die as 13900K/KS/KF, the same desktop chip used in mobile configuration. did you use AWCC "performance" mode or OC? those would accelerate the issue as well.

I think I had it in performance mode while gaming.

 

[greymaterial]

I think I had it in performance mode while gaming.

from what I see with my g-series 7630, the awcc performance mode would oc the cpu to 5.5GHz on two best binned p-cores and 5.3 on the rest, apart from the fan curve tuning. I only used the g button to max the fan speed as that one is without oc; manually i down tune the freq and power with XTU as well. see if you can tune down with XTU for stability as well, i do so just preventatively.

 

[Taslios]

It's been RMAed and the board had been replaced. This particular board was itself an RMA replacement, which I suspect was a refurbished board. We were able to rule out RAM since Dell did send a technician out to replace my RAM. Unfortunately the issues persisted. It could have been a GPU issue, but I found it interesting that my CPU was an i9 and some of the symptoms matched what the article mentioned. This issue was something that seemed to develop over time. When I got the laptop back, from the first RMA, I was able to game on it just fine. I took an extended break from said game then couldn't launch it anymore when I tried picking it back up.

I mean RMA the CPU with Intel. this sounds exactly what others have experienced as a CPU error... so very very flustering as the CPU is always the last thing and this really seems to be a memory/PCIe,/GPU issue when it is not.

 

[Taslios]

Part of the article just doesn't look right.

"with some institutions reportedly jumping to AMD’s Ryzen 9000 CPUs.

AMD has been having some issues on its side as well, with the launch of the Ryzen 9000 chips pushed back a couple of weeks due to an unspecified issue"

How can you be jumping to something that's not released yet?
Why would you be planing now to buy something that is already having problems without knowing it will be fixed properly?

Several Companies have said all "Future" replacements will be AMD 9000 series.

 

[Zerk2012]

Several Companies have said all "Future" replacements will be AMD 9000 series.

Thats fine I'm just saying not sure if I would jump on them as soon as its released.
I don't think AMD will ever say what they found that made them recall what was shiped already and delay the release. For all we know it could of been from the ones they have probably been testing full bore started having similar issues as Intel has from trying to push the bar to high. ( or not) I just like to wait a bit on new releases.

I've never seen a problem with a processor degrading as fast as the Intel chips look like they are.

I had a 2500K that did over a 4 year span but you can't complain when you have it @5.2 and then drop it to 5.0 and 4.8 in 4 years time.

 

[CelicaGT]

is your cpu i9-13900HX? this cpu has the same B0 die as 13900K/KS/KF, the same desktop chip used in mobile configuration. did you use AWCC "performance" mode or OC? those would accelerate the issue as well.

Intel claims laptop processors are not affected. We will see. As another owner of an i9 13900HX (Dell G15 5530) I am watching this all very closely. I have made sure my machine has all the OC features disabled in BIOS, and I nuked the factory Win11 install along with all the Dell/Alienware crapware. In doing so I lowered RAM usage, CPU utilization and overall system temperatures. If there is even a sniff that mobile chips are affected I'm applying for an RMA. The only real upside is that I got this machine at a deep discount*


*And now I know why...

 

[CelicaGT]

Thats fine I'm just saying not sure if I would jump on them as soon as its released.
I don't think AMD will ever say what they found that made them recall what was shiped already and delay the release. For all we know it could of been from the ones they have probably been testing full bore started having similar issues as Intel has from trying to push the bar to high. ( or not) I just like to wait a bit on new releases.

I've never seen a problem with a processor degrading as fast as the Intel chips look like they are.

I had a 2500K that did over a 4 year span but you can't complain when you have it @5.2 and then drop it to 5.0 and 4.8 in 4 years time.

Right? It's one thing if I OC it to the moon and it degrades, another thing entirely if it does it on its own.

 

[Taslios]

Thats fine I'm just saying not sure if I would jump on them as soon as its released.
I don't think AMD will ever say what they found that made them recall what was shiped already and delay the release. For all we know it could of been from the ones they have probably been testing full bore started having similar issues as Intel has from trying to push the bar to high. ( or not) I just like to wait a bit on new releases.

I've never seen a problem with a processor degrading as fast as the Intel chips look like they are.

I had a 2500K that did over a 4 year span but you can't complain when you have it @5.2 and then drop it to 5.0 and 4.8 in 4 years time.

Given the speed of the re-validation they are doing... I almost suspect it is something cosmetic like a misprint of the SN on the IHS..

According to a buddy who works at Microcenter, they have already gotten the first shipment of "re-validated" chips in and the only reason for the delay is that they would not have had the full stock by the 31st. .

 

[Zerk2012]

Right? It's one thing if I OC it to the moon and it degrades, another thing entirely if it does it on its own.

Well I think Intel did clock them to the moon or might of been further to Mars.

When it comes down to a pride thing to try to one up the competition if the speed (and the voltage needed to get it) for a bit of extra performance was 100% the reason it was a bad business decision.

Just to much to loose and really not much to gain. That would be hind sight since I'm sure they were not expecting these problems.
You have people that buy Intel, others that buy AMD, then the probably much smaller % of people that actually know enough to buy the best performance for the dollar, or unlimited money that just buys the best regardless of price.
So I couldn't really see the loss of market share being much of a issue if they were a % or 2 off the AMD chips.

 

[Zerk2012]

Given the speed of the re-validation they are doing... I almost suspect it is something cosmetic like a misprint of the SN on the IHS..

According to a buddy who works at Microcenter, they have already gotten the first shipment of "re-validated" chips in and the only reason for the delay is that they would not have had the full stock by the 31st. .

I would find the first part of that hard to believe. How would you already have them shipped and nobody in QC had seen some print wrong?

 

[CelicaGT]

Well I think Intel did clock them to the moon or might of been further to Mars.

When it comes down to a pride thing to try to one up the competition if the speed (and the voltage needed to get it) for a bit of extra performance was 100% the reason it was a bad business decision.

Just to much to loose and really not much to gain. That would be hind sight since I'm sure they were not expecting these problems.
You have people that buy Intel, others that buy AMD, then the probably much smaller % of people that actually know enough to buy the best performance for the dollar, or unlimited money that just buys the best regardless of price.
So I couldn't really see the loss of market share being much of a issue if they were a % or 2 off the AMD chips.

I have zero brand loyalty, I have had AMD and Intel systems in the past and have been an enthusiast for either. This particular purchase was based on the desire for a 4060 powered laptop (going 4070 bumped up to a 1440p display, erasing any framerate increase). I would have liked to try out an AMD offering just for kicks but Dell being Dell there was nothing in the line up that was desirable. In the case of my specific configuration, I could have gotten an i7 example with 16GB of RAM in single DIMM with the plan to upgrade the RAM to 16/32GB dual DIMM for around 100 to 150CAD. This unit, came with 32GB dual DIMM and the now questionable i9 for 100CAD less on sale than the prior (everything else the same) so I went that route. Time will tell if it was the right thing to do. I also admit, since my desktop is an AMD X3D part, to a strong curiosity to try out Intel's BIGlittle design.

(edit: to add) I just have to say, if these mobile i9's are affected I'm terribly disappointed. Not because of the processor but because this G15 is a nice little package. Wonderful 165Hz Gsync display, decent ass keyboard great overall build quality.

 

[JarredWaltonGPU]

I don't think AMD will ever say what they found that made them recall what was shiped already and delay the release. For all we know it could of been from the ones they have probably been testing full bore started having similar issues as Intel has from trying to push the bar to high. ( or not) I just like to wait a bit on new releases.

The most likely cause seems to be something to do with packaging, specifically with the desktop chips that have two or three die on the package (depending on the model). And it's probably not something that affected all of the chips, but some were impacted and whatever problem was identified was severe enough to have AMD pull everything back in. Wouldn't be surprising if the issue only impacts the two CCD models and not the single CCD variant, or vice versa. We do know (I think?) that the Ryzen AI chips aren't impacted, but those are single die monolithic chip designs that don't involve any more complex packaging steps and likely assembled on a different line.

I also want to know what AMD intends to do with all these first-round chips that have been recalled. Presumably a bunch will work fine, or they wouldn't have passed any level of quality control. But all of that quality testing is part of the main packaging, testing, etc. process and it's fully automated AFAIK. Pulling chips out of retail boxes to retest them just isn't part of the normal workflow, so there could be tens of thousands (or more?) of Ryzen 9000 chips that may or may not be okay.

I'm guessing AMD will probably end up selling them at a steep discount to some company that will in turn do all the unboxing, testing, etc. to weed out the faulty chips. Maybe a big OEM will do that, on the understanding that some percentage (5? 10? 25? Who knows!) will not work. But very likely AMD won't ever state definitively what happened and what was done to fix the issue. I mean, it could be as simple as one machine on one of the production lines having issues, out of however many machines are available, so that some set percentage of chips ended up bad. 🤷‍♂️

 

[CelicaGT]

The most likely cause seems to be something to do with packaging, specifically with the desktop chips that have two or three die on the package (depending on the model). And it's probably not something that affected all of the chips, but some were impacted and whatever problem was identified was severe enough to have AMD pull everything back in. Wouldn't be surprising if the issue only impacts the two CCD models and not the single CCD variant, or vice versa. We do know (I think?) that the Ryzen AI chips aren't impacted, but those are single die monolithic chip designs that don't involve any more complex packaging steps and likely assembled on a different line.

I also want to know what AMD intends to do with all these first-round chips that have been recalled. Presumably a bunch will work fine, or they wouldn't have passed any level of quality control. But all of that quality testing is part of the main packaging, testing, etc. process and it's fully automated AFAIK. Pulling chips out of retail boxes to retest them just isn't part of the normal workflow, so there could be tens of thousands (or more?) of Ryzen 9000 chips that may or may not be okay.

I'm guessing AMD will probably end up selling them at a steep discount to some company that will in turn do all the unboxing, testing, etc. to weed out the faulty chips. Maybe a big OEM will do that, on the understanding that some percentage (5? 10? 25? Who knows!) will not work. But very likely AMD won't ever state definitively what happened and what was done to fix the issue. I mean, it could be as simple as one machine on one of the production lines having issues, out of however many machines are available, so that some set percentage of chips ended up bad. 🤷‍♂️

They forgot the stickers....?


(/sarcasm if it wasn't obvious. I have skin in the Intel game fwiw and trying to make light of it.)

 

[Hotrod2go]

When they are pushing the limits from a physics point of view with 10nm node, hybrid architecture & high clocks speeds on the K models in order to compete with AMD's 5nm node Zen 4, Intel are living in some sort of dream world wishing the laws of physics would do as Intel wants it to do... not going to happen! & this is the result.