Question Weird artifacts, faulty cpu? BIOS update required?

Toggle sidebar Toggle sidebar
GoguDRks

GoguDRks

Aug 8, 2023
13
0
10
I recently bought a r5 5500 and a B550M ds3h and I've had some issues with them so far. Firstly, the cpu came with some bent pins, that I was able to straighten and fits perfectly into the socket. Secondly, my pc would not see more than 2 ram slots, although all of them were populated. Reseated the cup and screwed in the cooler more than I was comfortable with, which seems to have solved the issue.

But every time I am entering any browser, there are some weird artifacts showing up, and every single page is crashing, despite trying 3 browsers, despite the page I was loading (I'm getting crashes even on a blank page). The error message I'm receiving on chrome is "STATUS_ACCEES_VIOLATION" which I've read that it's due to a faulty cpu.

But what's even weirder is that under synthetic stress tests on the cpu and RAM, these artifacts disappear and all pages are loading, although a bit slower because the ram is 95% full and cpu at 100%. So, here are the things I am suspecting:
1. The cpu is indeed faulty
2. The bios needs to be updated
3. I might have to undevolt the cpu
4. The motherboard is faulty
5. The psu is not strong enough (very unlikely based on what I've seen so far)
6. I need to do something in the bios regarding the pcie version of the GPU, but I could not find that setting. I suppose the mobo sees the gpu pcie as a 4.0 and the cpu does not support 4.0 from what I'm aware of.

Specs on the pc I am having issues with:
Ryzen 5 5500
Rx 5600xt
B550M ds3h
Kingston nv2 1tb and kingston a400 (currently I have the a400 because that's the ssd that I used on the old build and I have some stuff to test)
Psu Segotep I believe 500w bronze 85% efficiency
4x16gb ram corsair vengeance 3200 cl16 (xmp not enabled, currently running at 2133)

Specs on the pc where everything works fine
I5 6500
Rx 5600xp
B150m hds
Kingston a400
Psu Segotep the one from the above
2x16gb hyperx fury 2133 (tested the corsair ones and were working fine)

Later updates: I can play some games but, for example, if I try playing Black Desert, the game simply crashes right when the terrain loads.

Ram does not seem faulty, works perfectly fine in the i5 system.

Cpu temps under full load for a long time are fine, staying at a strong 50°, peak was 63°
No issues appear within the synthetic test program (which is supposed to show any issues).

I've also tried the GPU in the 2nd pcie slot, same artifacts, but less frequent and I was actually able to load some pages.

I'll try attaching some photos of the artifacts in the comments.

Any suggestion is welcome, thanks!
 
I recently bought a r5 5500 and a B550M ds3h
What were you working with prior to these? If you migrated the OS drive from an old build without reinstalling the OS, you will need to reinstall the OS.

Psu Segotep I believe 500w bronze 85% efficiency
Got a link to the PSU? How old is the unit?

BIOS version for your motherboard? You should enable X.MP for your ram, mind you.
 
Lutfij said:
What were you working with prior to these?
Click to expand...
I5 6500, b150m hds, 2x8gb hyperx I don't remember something old 2133mhz, same psu, a400 kingston, 1tb wd hdd

Lutfij said:
If you migrated the OS drive from an old build without reinstalling the OS, you will need to reinstall the OS.
Click to expand...
I initially did a fresh install on the nv2 which went smoothly, but I plugged in the a400 as I had some apps on that one, that I wanted to test.
Same issue occurs on both SSDs, so I doubt there's an OS compatibility issue.
Windows 10 pro on both SSDs, latest updates installed.

Lutfij said:
Got a link to the PSU? How old is the unit?
Click to expand...
Not even segotep, but sirtec, it's that old that the name is not written on it anymore, but it gets the job done.
Idk how much you would understand, as it's in Romanian

Lutfij said:
BIOS version for your motherboard?
Click to expand...
The base one, did not try yet to update it, some letters, 3 in the amount

Lutfij said:
You should enable X.MP for your ram, mind you.
Click to expand...
I had it enabled but disabled it as I thought there was something to do with it. Either way, such artifacts should not show up regardless the XMP profile
 
PSU is indeed a really bad quality one, but I feel like your issues are not caused by it, right now. Though I'll strongly suggest you upgrade to a better unit in the near future.

On the Intel build are you using the same GPU as in the AMD one? Is the card working properly there?

Before anything, as the suggestions above, do the BIOS update and eventually remove the GPU drivers with DDU and reinstall.

If nothing changes I'd look at the part that was the problem in the first place, the CPU. If you had to actually force a lot of pressure on the CPU with your cooler, that leads me to believe that the pins were not straightened out completely/maybe the top of a pin is missing, leading to improper contact. Problem is that the metal becomes weaker and you might rip them off if you try to straighten them again. Another quick visual inspection on the pins wouldn't hurt though.

EDIT: Would be also worth trying to remove the RAM modules in the slots you said were not working before. Try running only with the 2 DIMMs that were working before tightening down the cooler to uncomfortable levels. If you're there loosen the CPU cooler a bit, since you won't run all DIMMs. See if that does anything.
 
Last edited:
ChumP said:
On the Intel build are you using the same GPU as in the AMD one?
Click to expand...
Yes, the exact same gpu and these artifacts are not showing up there. I've also tried the corsair ram on the Intel build and they seem to be working fine.
ChumP said:
remove the GPU drivers with DDU and reinstall.
Click to expand...
The issue is that the matter can be replicated on the Intel's OS (Kingston a400 has my old OS, which was used in the intel build, Kingstone NV2 is the new SSD which has the OS for the new build), which does not happen when I have the ssd in the intel build.
ChumP said:
If you had to actually force a lot of pressure on the CPU with your cooler
Click to expand...
I had to force more pressure on the CPU with the cooler more than I was comfortable with, but I initially screwed the cooler loosely, and that showed up in the temperatures, as they were constantly over 90 degrees under load, whilst right now they are low 50s.
ChumP said:
the pins were not straightened out completely/maybe the top of a pin is missing, leading to improper contact
Click to expand...
I've inspected the pins individually, from a lot of points, and I am more than certain that the top of the pins appear untouched, only maybe 1 or 2 pins have not been perfectly straighten up, but it is unnoticeable if you don't pay really close attention (I'm talking a difference of less than a 10th of a millimeter) but I'll give it another shot though.
ChumP said:
remove the RAM modules in the slots you said were not working before
Click to expand...
I've tried the RAM modules in every single possible configuration, which led me into the exact issue, the artifacts that I'm seeing on browsers, crashes etc. so I doubt there's an actual issue with the RAM.
ChumP said:
Try running only with the 2 DIMMs that were working before
Click to expand...
The 2 slots that were working before were not in dual channel, they were the A1 and A2 slots.

Edit: would undervolting the cpu solve this matter? For me, that seems to be the most logical thing and I don't want to try yet, as I've never done it.
 
AM4 and Corsair Vengeance don't mix well. Is there other RAM you can test with? If not, bump up DRAM voltage 0.05v and see if it's stable running a single stick.
 
kira-faye said:
Is there other RAM you can test with?
Click to expand...
Just tested with my other pair, hyperx fury 2133mhz, it seems to be reoccurring.
Bumped up the dram voltage by 0.1 (that's the only way it allowed me), no changes.
Bios updated to the latest version already, FDc.
Thanks, though!
 
COLGeek said:
Please explain this comment. Zero issues in the piles of AM4 systems I have built using Corsair (and even Vengeance versions) in all of them.

Specific models could be an issue, but as stated this is a rather broad comment.
Click to expand...
It's a trend I and others elsewhere have noticed. If someone makes a "does not POST" or "machine is unstable" thread and they're on AM4, chances are they're running a Corsair Vengeance kit.

I had problems with stability that were tracked down to RAM, went out and bought a 32gb (2x16) DDR4-3600 CL19 kit of Corsair Vengeance to replace it, but was still having occasional crashes. Through process of elimination, it was the RAM again. Raise voltage and it's a more stable but still not 100%. Return that kit, get some (harder to run at speed for the IMC) 64GB (2x32) G.Skill Trident Z DDR4-3600 CL18, works great, no XMP tweaking needed, no more stability problems.

People kept blaming the RAM and I was initially very defensive, but I've seen the pattern myself since.
 
Last edited:
kira-faye said:
It's a trend I and others elsewhere have noticed. If someone makes a "does not POST" or "machine is unstable" thread and they're on AM4, chances are they're running a Corsair Vengeance kit.
Click to expand...
Sources? Seriously, never experienced that. Just need to be careful with broad statements that could lead to folks spending unnecessary money to NOT solve a problem.

Assessing the specific model numbers used would be a more useful/informed activity. That "could" have merit.

Is this case, the OP changed brands altogether and the problem remains. Lets focus on that.
 
GoguDRks said:
Yes, the exact same gpu and these artifacts are not showing up there. I've also tried the corsair ram on the Intel build and they seem to be working fine.

The issue is that the matter can be replicated on the Intel's OS (Kingston a400 has my old OS, which was used in the intel build, Kingstone NV2 is the new SSD which has the OS for the new build), which does not happen when I have the ssd in the intel build.

I had to force more pressure on the CPU with the cooler more than I was comfortable with, but I initially screwed the cooler loosely, and that showed up in the temperatures, as they were constantly over 90 degrees under load, whilst right now they are low 50s.

I've inspected the pins individually, from a lot of points, and I am more than certain that the top of the pins appear untouched, only maybe 1 or 2 pins have not been perfectly straighten up, but it is unnoticeable if you don't pay really close attention (I'm talking a difference of less than a 10th of a millimeter) but I'll give it another shot though.

I've tried the RAM modules in every single possible configuration, which led me into the exact issue, the artifacts that I'm seeing on browsers, crashes etc. so I doubt there's an actual issue with the RAM.

The 2 slots that were working before were not in dual channel, they were the A1 and A2 slots.

Edit: would undervolting the cpu solve this matter? For me, that seems to be the most logical thing and I don't want to try yet, as I've never done it.
Click to expand...
Are you using the stock AMD CPU cooler (Wraith)? The version with 4 screws?

If so, those should be tightened until they won't turn any more. Seems tight, but necessary for optimal cooling.

Is your motherboard BIOS running the most recent version available for it?

Are you overclocking anything? If so, stop.
 
COLGeek said:
Sources? Seriously, never experienced that. Just need to be careful with broad statements that could lead to folks spending unnecessary money to NOT solve a problem.

Assessing the specific model numbers used would be a more useful/informed activity. That "could" have merit.

Is this case, the OP changed brands altogether and the problem remains. Lets focus on that.
Click to expand...
Typically I'd agree that limiting the advice to a specific kit or range is important, but Corsair's tendency to change ICs and keep using the same SKU means there's no easy way to do so here.
 
kira-faye said:
Typically I'd agree that limiting the advice to a specific kit or range is important, but Corsair's tendency to change ICs and keep using the same SKU means there's no easy way to do so here.
Click to expand...
That is true of other brands as well. You make a valid point in regards to mixing/matching memory. Matched sets (sold that way) should always be used for optimal performance. Users must not rely on a product line name.
 
COLGeek said:
If so, those should be tightened until they won't turn any more.
Click to expand...
Yup, that's exactly what I did, and screwing them until they wont turn made me feel a bit uncomfortable, as they seemed too tight.
COLGeek said:
Is your motherboard BIOS running the most recent version available for it?
Click to expand...
Yep, latest BIOS update, FDc, released Sep 20, 2023.
COLGeek said:
Are you overclocking anything?
Click to expand...
Everything is under the base settings, nothing has been overclocked so far, except enabling XMP, which I've disabled eventually.

To put the problem into perspective, I've attached a link on the first comment, with the artifacts I'm referring to.
They completely disappear when I'm running synthetic stress tests on the cpu, but I'm unable to verify the capabilities further, as the cpu is fully loaded. I.e. I can't open games to see if they are loading up.
 
On the AMD build, I assume the CPU was used before you bought it (bent pins). What else was used before in that build (not new when you installed)?

Also, if you booted this system from a Live Linux distro, do the issues occur there as well?
 
COLGeek said:
What else was used before in that build
Click to expand...
Everything has been used before (concretely, bought 2x2 2x16GB Corsair Vengeance kits, the Ryzen 5 5500, the mobo and the M2 SSD), I bought them from a known retailer from Romania, as "Resealed", and I've though that the CPU/mobo might have been damaged, but I assume the system should not have booted in the first place, or it should not have been stable at all under full load, whilst, under full load, it is way more stable than under no load.
I'm a bit tech savvy myself and I know what is supposed to happen and what not, but I've never had the possibility to overclock/underclock/undervolt a CPU, as the intel CPU is locked

As for your 2nd question, I tried my best to stay away from Linux 🙁(( but I guess I'll have to give it a try.

Update: An important thing to be mentioned I suppose is that, on the HWMonitor, the GPU clocks appear to be very low, and the artifacts that are showing up seem to be due to the GPU. Concretely, I have the following clocks, which, for me, seem to be very low even for no load:
Graphics: between 6Mhz - 36MHz
Memory: 200MHz
Processor: between 50-150MHz

Update 2: I've just tried an old GT730 I believe and it's even worse.
 
Last edited:
Update 3, I might have solved the matter.
So, what I've done is that I tried putting in all of the ram dimms into the mobo, to try enabling xmp on all of them, but seems like one of the slots might be damaged (which doesn't make sense as it used to work, most likely the contact of the pins was imperfect, I was stuck at the mobo/bios logo page), unplugged all of them and let only one dimm. It did not solve the matter, so I've thought maybe undervolting and underclocking the cpu would work. It did not, of course!

I've installed the OC tweaker from the mobo app, as I wanted to undervolt it a bit more (I was initially at 3.2 ghz with 1.070V), but saw the overclock button and my inner thoughts won. Pressed it, cpu performance went to 4.5 and all my issues appear to have been solved.

I'm currently installing the game I was having issues with (black desert) to see whether or not that was the issue.

But something still remains unclear to me, the bios settings have been restored, for cpu at least (it went back to 3.6 base clock and 4.25 oc clock) and I have not overclocked it again from the app and, for some reason, the artifacts are not showing up, nor the pages are crashing anymore.

I'll update y'all as soon as I have the game up and running, to let y'all know if the temporary overclock solved the problem, maybe you might help someone else with similar issues.

Thanks a lot for the implication, though, really appreciate it!
 
Last edited:
Update 4: All good, the temporary OC seems to have solved the matter. Currently running under base specs, no oc applied, no XMP, but with XMP enabled it works fine as well.
 
Yess, everything works as expected, I'm not certain what was causing the problem exactly but it was certainly the cpu.
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom