Discussion Zen 5 Builders thread

Page 39 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Dave3000

Golden Member
Jan 10, 2011
1,491
108
106
I just saw this article pop up while browsing the web and this got be really concerned about my recent purchase:


I recently upgraded to a Ryzen 9800X3D with an MSI X670E Tomahawk motherboard for my main system and passed down my Ryzen 7800X3D, also with an MSI X670E Tomahawk motherboard, to my secondary system which was formerly a pre-built system with an i5-12400 and a crappy H610M motherboard, crappy CPU cooler, and crappy case (however, it looks sleek but that's the only good thing about that case IMO). I wonder if I should just return the 9800X3D and new motherboard and go back to my 7800X3D for my main system and i5-12400 for my secondary system with the new case, new motherboard, and new CPU cooler. I already have the new CPU cooler and new case. I was planning on reselling my i5-12400 pre-built and having the 7800X3D system take it's spot but now I'm not so sure. On my 9800X3D system I don't run EXPO settings, only stock, that includes JEDEC DDR5-5600 (default) for my 9800X3D.
 
Last edited:

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
27,023
15,963
136
I just saw this article pop up while browsing the web:


I recently upgraded to a Ryzen 9800X3D with an MSI X670E Tomahawk motherboard for my main system and passed down my Ryzen 7800X3D, also with an MSI X670E Tomahawk motherboard, to my secondary system which was formerly a pre-built system with an i5-12400 and a crappy H610M motherboard, crappy CPU cooler, and crappy case. I wonder if I should just return the 9800X3D and new motherboard and go back to my 7800X3D for my main system and i5-12400 for my secondary system with the new case, new motherboard, and new CPU cooler. I already have the new CPU cooler and new case. I was planning on reselling my i5-12400 pre-built and having the 7800X3D system take it's spot but now I'm not so sure. On my 9800X3D system I don't run EXPO settings, only stock, than included JEDEC DDR5-5600 for my 9800X3D.
That article is click bait IMO, as NO EXAMPLES of dead cpus were reported. Just motherboard that had out of AMD spec high voltages. (correct me if I missed the examples)
 
  • Like
Reactions: Makaveli

Hail The Brain Slug

Diamond Member
Oct 10, 2005
3,766
3,070
146
That article is click bait IMO, as NO EXAMPLES of dead cpus were reported. Just motherboard that had out of AMD spec high voltages. (correct me if I missed the examples)
A handful of examples show up online and people assume the entire platform is cursed. I think it needs to be made clear there are only double digit reports of any issues similar to this whatsoever, if even that (citing the reddit thread where they were attempting to track all known instances of issues like this).

Meanwhile the Intel problem had individuals that would go through 2 or 3, sometimes more dead intel CPU's, RMAing each time for a new processor.

I personally know more than 3 people that had 2 or more intel cpu's die on them. The internet collectively only knows of a handful of dead AM5 processors. These incidence rates are several orders of magnitude disparate.
 

Makaveli

Diamond Member
Feb 8, 2002
4,935
1,520
136
I just saw this article pop up while browsing the web and this got be really concerned about my recent purchase:


I recently upgraded to a Ryzen 9800X3D with an MSI X670E Tomahawk motherboard for my main system and passed down my Ryzen 7800X3D, also with an MSI X670E Tomahawk motherboard, to my secondary system which was formerly a pre-built system with an i5-12400 and a crappy H610M motherboard, crappy CPU cooler, and crappy case (however, it looks sleek but that's the only good thing about that case IMO). I wonder if I should just return the 9800X3D and new motherboard and go back to my 7800X3D for my main system and i5-12400 for my secondary system with the new case, new motherboard, and new CPU cooler. I already have the new CPU cooler and new case. I was planning on reselling my i5-12400 pre-built and having the 7800X3D system take it's spot but now I'm not so sure. On my 9800X3D system I don't run EXPO settings, only stock, that includes JEDEC DDR5-5600 (default) for my 9800X3D.
unless you are using an asrock motherboard I wouldn't bother.

Zero issues on my Asus board.
 
Jul 27, 2020
24,583
17,089
146
Update from ASROCK themselves at Computex:


Summary:

The video discusses issues surrounding ASRock motherboards and Ryzen 9000 series CPUs, notably premature CPU failures linked to aggressive BIOS settings. The creator visited ASRock's booth at Computex, where they learned that recent BIOS updates aim to resolve these problems.
Key details include:
  • The problem was tied to the PBO (Precision Boost Overdrive) settings, specifically the ED and TDC current limits, which were set too aggressively on mid-range and high-end motherboards (e.g., B650E, X670E Taichi, B850 Steel Legend). These settings, combined with AMD's imposed limits, caused CPUs—especially early samples of Ryzen 9000 series—to die prematurely.
  • ASRock claimed that their latest BIOS updates have adjusted these PBO settings and shadow voltages to safer levels, making CPUs run more reliably and preventing further failures.
  • Testing with a Ryzen 7 9800X3D showed that after BIOS updates, performance metrics (cinebench scores, power consumption, FPS) remained stable, though FPS was slightly lower, indicating more conservative operation.
  • The issue appears to originate from the CPUs being pushed beyond their thermal and power limits, especially with high-performance cooling like water coolers. ASRock's new BIOS versions have dialed down the PBO settings to mitigate this.
  • There is some ambiguity about whether AMD's original limits were too high or if motherboard manufacturers were pushing CPUs beyond safe thresholds. ASRock indicated that their settings now align with other manufacturers (MSI, ASUS, Gigabyte), and they emphasized that the problem was mainly due to overly aggressive PBO configurations.
  • The creator emphasized the importance of stability over maximum performance, criticizing the industry trend of shipping CPUs with aggressive overclocking out of the box, which can jeopardize hardware longevity.
  • They suggested that motherboard manufacturers could offer a "stability mode" during first setup, allowing users to choose safer preset configurations rather than defaulting to performance-optimized but risky settings.
  • The issue raises concerns about the long-term reliability of early Ryzen 9000 series CPUs, especially for users with older batches or weaker components, and whether extended warranties or replacements might be necessary.
  • The creator called for more transparency from ASRock and AMD, recommending joint communication to clarify the root cause, fixes, and and responsible parties.
  • Personal notes include ongoing testing with a Ryzen 9 9950X, awaiting AMD's response for a replacement, and noting that a Ryzen 7950X has been stable so far.
  • Overall, the situation underscores the need for industry-wide emphasis on stability and user safety, rather than solely maximizing out-of-the-box benchmarks, to prevent hardware failures and ensure user confidence.
 
  • Like
Reactions: Makaveli

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
27,023
15,963
136
I have multiple X670E Taichi board with 7950x and 7950x3d and one 9950x. None have failed, and I run them 24/7/365 at 100% load. There is something else missing here. All with the default shipped bios, and no updates.
 
  • Like
Reactions: igor_kavinski
Jul 27, 2020
24,583
17,089
146
They do say "early" Ryzen 9000 samples so maybe there's a problem with the chips themselves that was subsequently fixed in later steppings. Could be something as simple as changing resistor values or something like that.
 

Makaveli

Diamond Member
Feb 8, 2002
4,935
1,520
136
I have multiple X670E Taichi board with 7950x and 7950x3d and one 9950x. None have failed, and I run them 24/7/365 at 100% load. There is something else missing here. All with the default shipped bios, and no updates.
Its really hard to tell sometimes by just looking at that asrock reddit thread. There are still daily post of people burning cpu's but i'm starting to think its also user error.

  1. Some users using scaler x10 and clearly have no idea what they are doing
  2. Some of them are just copying other people settings and guide's from the net still don't know what they are doing
  3. When you ask them if they manually set vsoc the response is what is that
  4. Don't know how to PBO tune
  5. alot of first time builders etc
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
27,023
15,963
136
You don't use PBO, do you?

Its really hard to tell sometimes by just looking at that asrock reddit thread. There are still daily post of people burning cpu's but i'm starting to think its also user error.

  1. Some users using scaler x10 and clearly have no idea what they are doing
  2. Some of them are just copying other people settings and guide's from the net still don't know what they are doing
  3. When you ask them if they manually set vsoc the response is what is that
  4. Don't know how to PBO tune
  5. alot of first time builders etc
Real OC'ers can kill it for everyone unless they are expert, like detx0. I use PBO only to DETUNE a little, -30 on the one setting, and the temp down to 85c.
 

gdansk

Diamond Member
Feb 8, 2011
4,121
6,828
136
I have an oddity to report.
For 2.5 years I have had the same motherboard (ASRock X670E Steel Legend) and haven't had any USB issues.
But now, after changing my memory from 2x32GB CL36 to 2x48GB CL30, I have USB disconnects all the time. I ran memory test all weekend without issues. I suppose then that memory itself isn't defective.

Do I have to revert to my older, slower memory? I'm not sure why switching to CL30 memory would cause this. No other devices were changed when this issue started appearing. The CPU was upgraded to a 9950X3D in March but the USB disconnects only appeared after I upgraded the memory a week ago.
 

Makaveli

Diamond Member
Feb 8, 2002
4,935
1,520
136
I have an oddity to report.
For 2.5 years I have had the same motherboard (ASRock X670E Steel Legend) and haven't had any USB issues.
But now, after changing my memory from 2x32GB CL36 to 2x48GB CL30, I have USB disconnects all the time. I ran memory test all weekend without issues. I suppose then that memory itself isn't defective.

Do I have to revert to my older, slower memory? I'm not sure why switching to CL30 memory would cause this. No other devices were changed when this issue started appearing. The CPU was upgraded to a 9950X3D in March but the USB disconnects only appeared after I upgraded the memory a week ago.
Is it the same bios version or has that changed?

I know USB disconnects was a big issue on AM4 your post is the first i've heard of it on AM5.

For troubleshooting after the memory change did you reset bios to default settings?
 

biostud

Lifer
Feb 27, 2003
19,580
6,647
136
Is it the same bios version or has that changed?

I know USB disconnects was a big issue on AM4 your post is the first i've heard of it on AM5.

For troubleshooting after the memory change did you reset bios to default settings?
Sometimes I have to unplug my keyboard and re-plug it to to power on. Maybe 1/20 times I turn on my computer.
 

DrMrLordX

Lifer
Apr 27, 2000
22,575
12,441
136
Real OC'ers can kill it for everyone unless they are expert, like detx0. I use PBO only to DETUNE a little, -30 on the one setting, and the temp down to 85c.
The point being, at least according to ASRock, the failures they've documented have all been related to people using PBO and the old UEFI/microcode revs supplying unsafe voltages under those circumstances. Hence the updates. If you personally aren't using PBO to increase performance, then you wouldn't be at risk (assuming what ASRock says is true).
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
27,023
15,963
136
The point being, at least according to ASRock, the failures they've documented have all been related to people using PBO and the old UEFI/microcode revs supplying unsafe voltages under those circumstances. Hence the updates. If you personally aren't using PBO to increase performance, then you wouldn't be at risk (assuming what ASRock says is true).
If only overclockers are the ONLY ones causing failures, why are we blaming asrock and why is this an issue that is on the web ?
 
Jul 27, 2020
24,583
17,089
146
AMD may need to integrate PBO testing in their validation workflow going forward and any chips that appear to be weak will need to be pushed as Ryzen PRO CPUs.