News Some Ryzen 7 9800X3D CPUs are allegedly dying prematurely — over 100 cases documented based on user feedback

Toggle sidebar Toggle sidebar
Is there any possibility of these cases being user errors?

I don't think such incidents should discourage anyone wishing to buy 9800X3D.

After all, hardware failures happen all the time, with all sorts of products.
 
  • Like
Reactions: Makaveli
Hopefully someone is actually investigating over at AMD. Even if it's not their fault they need to figure out what is going on. Bad chips do happen and it could just be as simple as that, but if I'm remembering the reports right one person had two die.
 
  • Like
Reactions: jlake3
valthuer said:
Is there any possibility of these cases being user errors?

I don't think such incidents should discourage anyone wishing to buy 9800X3D.

After all, hardware failures happen all the time, with all sorts of products.
Click to expand...
I think alot of this is user error. With the volume of X3D processors that have been sold and when I look at the reddit post.

You have tons of people building that have zero experience and also alot of first time AMD users.

I've had zero issues with my 9800X3D on my Asus board, but i'm also someone with a ton of experience building pc's been doing this since the early 90s.
 
  • Like
Reactions: Hotrod2go and valthuer
From the preliminary data it looks like some sort of asrock bios settings maybe lacking some safeguard so some random combinations might overvolt it periodically again? Or maybe the flipped 3D vcache could still have some hidden degradation issues. I wonder how many of the burnt out occurred using PBO or not also.
 
  • Like
Reactions: rluker5 and LabRat 891
Makaveli said:
I think alot of this is user error. With the volume of X3D processors that have been sold and when I look at the reddit post.

You have tons of people building that have zero experience and also alot of first time AMD users.

I've had zero issues with my 9800X3D on my Asus board, but i'm also someone with a ton of experience building pc's been doing this since the early 90s.
Click to expand...

Yeah, stuff like this will always occur.

And it's not easy for someone to admit to a layman's mistake.

Remember the Melt Gate with 4090/5090?

People make a big deal out of anything these days.

P.S. Apart from doing simple things, like changing a GPU, placing a new SSD, or changing RAM modules, i have zero experience from PC building, so i prefer to trust my local PC service stores.

I would hate to do something that will cause permanent damage to my rig.
 
  • Like
Reactions: Makaveli
There were about 2 or 3 user errors where they crunched the socket because they built the PC upright (mobo in vertical orientation when installing CPU).

FYI, you're supposed to install the CPU onto the mobo first and the mobo should be laying flat on a tabletop. Then you install the mobo into the case. IDK if the user manual tells you this, but no one bothers to read these days.
 
  • Like
Reactions: Hotrod2go and bit_user
My 5600 was part of a bad batch of early 5000series chips. Little less than a year after I got it, it lost a core.

As long as the end-user isn't directly responsible from mis-use, I'd hope a mere 1-2week RMA turnaround is all these customers had to deal with.

Still, we need to know if this is a 'diceroll' issue, or if ASRock (and others) have yet another firmware glitch cooking CPUs
 
valthuer said:
I seriously doubt it's happening at a percentage that would indicate a worrisome trend.

That's my point.

If i had to guess, the same goes for 9800X3D.
Click to expand...
Given, the great many 9800X3D (and other AM5 CPUs) sold, the mere 100+ cases is well sub-single-percentage.
This easily could be a 'defect batch' of CPUs, or a shared component all these board manufacturers use (voltage reg, etc. that itself is part of a 'bad batch').

It's happening enough, with enough consistency to not excuse as 'normal failure rate', but there's no (current) indicator that this is a far-reaching or commonplace "systemic" issue.
 
  • Like
Reactions: jlake3
My first 1800X: Died in a few months (like many early adopters)

My 5950X: Won't boost worth anything (nowhere near what TomsHardware, Techpowerup, and others saw under same loads) in both single and multi thread.


After 20 years of using only AMD my next CPU will be Intel.
 
And how many of those reports are real? Would be easy for some Intel diehard fans to go on reddit and say "my 9800x3d just died". We are not living in a magical world with rainbows and unicorns, and random people online can't be blindly trusted.
 
Man, I thought when I built my machine a couple years ago it was one of the worst times to have my machine die and have to replace it.

But it has certainly gotten worse. Much worse it seems.

The prices, the lack of supply, the problems seem to be never ending this generation. I certainly hope next generation turns out far better than this one.
 
YSCCC said:
From the preliminary data it looks like some sort of asrock bios settings maybe lacking some safeguard so some random combinations might overvolt it periodically again? Or maybe the flipped 3D vcache could still have some hidden degradation issues. I wonder how many of the burnt out occurred using PBO or not also.
Click to expand...
Seeing as how we have seen periodic overvolting at "stock" settings be a problem in the recent past with both Intel and AMD it certainly seems plausible. And not just core volts, all sorts of components could fall victim. Those last couple hundred MHz should be left up to the user.
 
Alvar Miles Udell said:
My first 1800X: Died in a few months (like many early adopters)

My 5950X: Won't boost worth anything (nowhere near what TomsHardware, Techpowerup, and others saw under same loads) in both single and multi thread.


After 20 years of using only AMD my next CPU will be Intel.
Click to expand...
Is your flair correct, and are you running four sticks of RAM @3600? You'll kill that CPU; it's rated for 2667 over four sticks. Incidentally, overclocking the RAM juices voltages and that will eat into the CPU's power budget, ergo, lower boost clocks.
 
  • Like
Reactions: tamalero
There are 3 failure types.
Infant mortality, often taken care of by burn in testing. Some parts will pass and fail soon after.
Old age, the part is worn out.
The third is failing within the normal lifetime of the part, the flat part of the bathtub curve. This may be caused by a weakness in the manufacture of the part with no external (to the part) influence or it may be caused by using the part outside of its rated limits.

100 ish parts in the scheme of things is less than a drop in the ocean but still warrants investigation. If there is a common cause it may be mitigated in software and it can be avoided in subsequent designs.
Commodity parts are made to the lowest cost for the feature set and expected lifetime.
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom