AMD Zen Discussion Thread

z0oT

.G. wrote: »

His GPU came from my system when I upgraded, the perks he gets from me being a tech head

It was fine when I had it granted that doesn't mean it still is. Stress testing is ongoing right now, hour and a half in with no problems and max temp 78c on the CPU and 75c on the GPU.

If it makes you feel any better - here's some anguish I've had in the last few hours. :pac:

I picked up an R5 2600 for my server as an upgrade from the R3 2200G I currently have in it. A bit over €100 - a steal. It should have been just a simple swap, pull out the 2200G, and stick in the 2600 & Windows Server would just boot up as before, but nothing is ever easy. :pac:

The system hung a little upon startup, but that's to be expected with new hardware. Windows booted and here I am thinking great! But then after a few seconds I get a BSOD (Critical Process Died). Reboot and after that I can't get into the BIOS anymore.

So I try removing the battery to clear the CMOS (only way to do it on the board I have), I get into the BIOS, then it freezes upon disabling CSM.

Reboot - no video, nothing. Clear the CMOS again pull out everything from the board except 1 DIMM & the CPU fan, get into the BIOS once more upon a reboot, flash to an older version using the IPMI, no video upon boot once again. Re-Flash to a newer version, reboot - this time I get video but it hangs at initializing once again.

So after going around in circles for a while, in an act of desperation, I put the 2200G back in once more and everything boots up just as it was before. After plugging everything back into the board everything is back to normal - all my network drives, services - everything.

Then I think was it bad contacting with the CPU socket - so I stick back in the 2600 once more, hangups & freezes again. Replace it with the 2200G once more and everything is fine again.

So now I'm wondering did I get a bad CPU. It'll be the 1st time in my 14 or so years at this madness that it'll have happened to me. I still have my old B450 motherboard, so I might give the new R5 2600 a shot in that to test it out there.

Either way I might just return it anyway - Seriously disappointing!

.G.

z0oT wrote: »

If it makes you feel any better - here's some anguish I've had in the last few hours. :pac:

I picked up an R5 2600 for my server as an upgrade from the R3 2200G I currently have in it. A bit over €100 - a steal. It should have been just a simple swap, pull out the 2200G, and stick in the 2600 & Windows Server would just boot up as before, but nothing is ever easy. :pac:

The system hung a little upon startup, but that's to be expected with new hardware. Windows booted and here I am thinking great! But then after a few seconds I get a BSOD (Critical Process Died). Reboot and after that I can't get into the BIOS anymore.

So I try removing the battery to clear the CMOS (only way to do it on the board I have), I get into the BIOS, then it freezes upon disabling CSM.

Reboot - no video, nothing. Clear the CMOS again pull out everything from the board except 1 DIMM & the CPU fan, get into the BIOS once more upon a reboot, flash to an older version using the IPMI, no video upon boot once again. Re-Flash to a newer version, reboot - this time I get video but it hangs at initializing once again.

So after going around in circles for a while, in an act of desperation, I put the 2200G back in once more and everything boots up just as it was before. After plugging everything back into the board everything is back to normal - all my network drives, services - everything.

Then I think was it bad contacting with the CPU socket - so I stick back in the 2600 once more, hangups & freezes again. Replace it with the 2200G once more and everything is fine again.

So now I'm wondering did I get a bad CPU. It'll be the 1st time in my 14 or so years at this madness that it'll have happened to me. I still have my old B450 motherboard, so I might give the new R5 2600 a shot in that to test it out there.

Either way I might just return it anyway - Seriously disappointing!

This stuff can be frustrating as hell! I feel your pain.

As an update on my end we ran aida64 for two hours with no issues. Within 5 minutes of ending the test he was un installing one of the ubisoft games, he hit cancel while it was doing it cos he thought it might delete his game saves and he got instant BSOD.

Eventually finished the process properly, have uplay and The Division 2 now installed on his ssd rather than the hdd. No BSOD yet but one game crash.

Kimbot

.G. wrote: »

This stuff can be frustrating as hell! I feel your pain.

As an update on my end we ran aida64 for two hours with no issues. Within 5 minutes of ending the test he was un installing one of the ubisoft games, he hit cancel while it was doing it cos he thought it might delete his game saves and he got instant BSOD.

Eventually finished the process properly, have uplay and The Division 2 now installed on his ssd rather than the hdd. No BSOD yet but one game crash.

Are the hdds brand new?

H3llR4iser

Kimbot wrote: »

Are the hdds brand new?

curious as well about that - this sounds more and more like a data I/O issue...or something so weirdly wrong with the CPU that only shows up during specific operations.

.G.

not brand new no but also came from my machine and were fine in it.

Kimbot

.G. wrote: »

not brand new no but also came from my machine and were fine in it.

I'm leaning towards one of the hdds been corrupt in a weird way, you had the OS on 1 and data on the other. I would go with a complete reformat of 2 drives.
Then as a test I would use 1 drive and put the OS, drivers, uplay and 1 game that mostly causes the bsod on it. When all is up and config turn it all off and do the exact same thing with second drive so they are both identical. Then go and test 1 drive at a time and see if either cause the bsods but you will probably need to put a good bit of testing time into each setup over 2/3 days and document the results and see how it looks.

z0oT

z0oT wrote: »

If it makes you feel any better - here's some anguish I've had in the last few hours. :pac:

I picked up an R5 2600 for my server as an upgrade from the R3 2200G I currently have in it. A bit over €100 - a steal. It should have been just a simple swap, pull out the 2200G, and stick in the 2600 & Windows Server would just boot up as before, but nothing is ever easy. :pac:

The system hung a little upon startup, but that's to be expected with new hardware. Windows booted and here I am thinking great! But then after a few seconds I get a BSOD (Critical Process Died). Reboot and after that I can't get into the BIOS anymore.

So I try removing the battery to clear the CMOS (only way to do it on the board I have), I get into the BIOS, then it freezes upon disabling CSM.

Reboot - no video, nothing. Clear the CMOS again pull out everything from the board except 1 DIMM & the CPU fan, get into the BIOS once more upon a reboot, flash to an older version using the IPMI, no video upon boot once again. Re-Flash to a newer version, reboot - this time I get video but it hangs at initializing once again.

So after going around in circles for a while, in an act of desperation, I put the 2200G back in once more and everything boots up just as it was before. After plugging everything back into the board everything is back to normal - all my network drives, services - everything.

Then I think was it bad contacting with the CPU socket - so I stick back in the 2600 once more, hangups & freezes again. Replace it with the 2200G once more and everything is fine again.

So now I'm wondering did I get a bad CPU. It'll be the 1st time in my 14 or so years at this madness that it'll have happened to me. I still have my old B450 motherboard, so I might give the new R5 2600 a shot in that to test it out there.

Either way I might just return it anyway - Seriously disappointing!

Figured out my issue.

With the R5 2600 installed, I swapped the Crucial DIMMs for a Corsair Vengeance LPX DIMM from my gaming desktop - and everything worked just fine. BIOS loads just fine, Windows boots & all my services & network drives are fine once again.

So it would seem the issue is the Crucial memory & R5 2600 pair. That memory has been a bit funny in the past. My old i7-6700k for instance would freeze in the BIOS if XMP was enabled with it.

A successful few hours but I guess I need new memory now.

.G.

Kimbot wrote: »

I'm leaning towards one of the hdds been corrupt in a weird way, you had the OS on 1 and data on the other. I would go with a complete reformat of 2 drives.
Then as a test I would use 1 drive and put the OS, drivers, uplay and 1 game that mostly causes the bsod on it. When all is up and config turn it all off and do the exact same thing with second drive so they are both identical. Then go and test 1 drive at a time and see if either cause the bsods but you will probably need to put a good bit of testing time into each setup over 2/3 days and document the results and see how it looks.

I've got two spare HDD's and one spare SSD I've tried all of them at different points in this process and it happens with all of them, they can't all be corrupt imo. The process has been one SSD for windows matched with one of the HDDs. So I've tried his SSD with his HDD, I've tried my spare SSD with both my HDDs one at a time and then his one. Have tried different sata cables too.

Both the SSD's are only 128GB so can't put anymore than windows and the Division 2 on it. He still hasn't had a BSOD since we put uplay and the one uplay game on the SSD but the game itself still crashes. He's playing steam games today with no issues so far and they are on the HDD.

It really seems to be some weird quirk between his particular system and uplay. Granted I bet I've jinxed it now and the steam games will crash:D

.G.

Just came across a decent deal for a 1TB M2 drive so I've bought that. Will get rid of all the other drives and just use that when it comes and do lots and lots of praying

jonerkinsella

z0oT wrote: »

Figured out my issue.

With the R5 2600 installed, I swapped the Crucial DIMMs for a Corsair Vengeance LPX DIMM from my gaming desktop - and everything worked just fine. BIOS loads just fine, Windows boots & all my services & network drives are fine once again.

So it would seem the issue is the Crucial memory & R5 2600 pair. That memory has been a bit funny in the past. My old i7-6700k for instance would freeze in the BIOS if XMP was enabled with it.

A successful few hours but I guess I need new memory now.

Send the modules back to Crucial under warranty. There good to deal with.

.G.

.G. wrote: »

Just came across a decent deal for a 1TB M2 drive so I've bought that. Will get rid of all the other drives and just use that when it comes and do lots and lots of praying

M2 just arrived so will find out soon although he's only had one BSOD since we moved Uplay and the one Uplay game that would fit onto his 128GB SSD drive. Still game crashes aplenty though of all games on all drives.

.G.

M2 installed, all other drives removed, yet another clean install of windows, game crashes continue unabated. At this stage I'm ready to abandon AMD forever. Never had such hassle with a PC than this one. Now on a new drive on a new Mobo with good ram, good PSU and GPU ( I believe) and still it goes on. :mad::mad: No BSODs though!

wotzgoingon

.G. wrote: »

M2 installed, all other drives removed, yet another clean install of windows, game crashes continue unabated. At this stage I'm ready to abandon AMD forever. Never had such hassle with a PC than this one. Now on a new drive on a new Mobo with good ram, good PSU and GPU ( I believe) and still it goes on. :mad::mad: No BSODs though!

I don't know what's going on with your system. Should have gone X570.

.G.

At this stage I don't think it'd matter. Surely now we are at the point where we've proved neither mobo is faulty, even the one gone back since the same issues continue.

The 3600 is the prime suspect at at this stage.

.G.

I'll put my psu and my 2080ti in his machine. After that all that is left is the cpu. At least then I'll have a reason to request an rma from Overclockers for it.

Mr Crispy

Sorry G, haven't read all your posts re: this issue, so you may have mentioned this already. But have you had a look at the pins on your CPU for any that might be damaged or missing?

.G.

Yeah I have I've taken it out several times. All good.

H3llR4iser

How did the 2080ti + PSU experiment go?

.G.

Haven't got round to it yet. In the meantime I've told him to drop graphics settings to medium or something and see what happens.

Cuddlesworth

.G. wrote: »

Haven't got round to it yet. In the meantime I've told him to drop graphics settings to medium or something and see what happens.

Sounds like it would have been easier to learn how to install a linux live OS and test the ram with a proper stress test. Lots of people moved away from memtest a long time ago. Doesn't put heat and power into the cpu and ram, it's not really a stress test anymore.

.G.

Like this? https://www.tecmint.com/install-linux-os-on-usb-drive/

Cuddlesworth

Yep, boot off the memory key, open a terminal window
run the command, put in the Sudo/root password when prompted.

"sudo yum install stressapptest"

Then this test for example would stresses the CPU, ram and cache at the same time. You have a 16 thread proc, your going to get 16 simultaneous read/write operations to memory.

"stressapptest -W --cc_test" You might need a "sudo" in front of this, been a while.

Memtest checks each chunk of memory for errors but its doesn't really stress the delivery mechanism at all. App's like prime95 really only stress the CPU. But there is a whole heap of stuff between CPU and the memory that get ignored. Hence why Google ended up creating that app.

harvindhillon

I was getting BSODs before with a faulty PSU, I would recommend testing with another one if you have handy.

I've been a pioneer with AMD Zen processors, started with the 1700x to 2700x and now 3900x and thus far apart from RAM issues with the 1700x in the early days, it has been really smooth experience.

Also I stupid question but I suppose you have the latest BIOS for your board?

.G.

Yeah latest BIOS. Haven't had a BSOD for a couple of weeks now to be fair, just game crashes. Ubisoft to be fair to them are still trying to help. They keep asking so DxDiag files and Msinfo files. They said they can see from these that Nvidia errors are popping up a fair bit and also windows errors which is perplexing since I'm on about the 5th clean install of windows at this stage which also means the Nvidia driver is a clean install too. Their support person also mentioned the PSU as being right on the edge of being able to handle a 3600 and a 1080 while fully loaded but any PSU calculator I've tried said it should be fine.

Only other one I have anyway is my own and I'm going to try that along with my own GPU at the weekend.

harvindhillon

I used to run a 1700x with a 1080, this was with a 450watt PSU. As soon as I upgraded to a 2700x and a 1080ti, my games crashes out of desktop quite often or I have a PC shutdown.

PSU was the problem, upgraded to a 650watt and it solved the problems.

I would still give it a try with a different higher power PSU just to be safe and rule it out, cause apart from GPU and the PSU, you've ruled everything else out.

.G.

Cuddlesworth wrote: »

Sounds like it would have been easier to learn how to install a linux live OS and test the ram with a proper stress test. Lots of people moved away from memtest a long time ago. Doesn't put heat and power into the cpu and ram, it's not really a stress test anymore.

OK so finally got around to this. Installed unbuntu on a spare SSD. ran the stress test for 2 hours. No errors. From what I read online, 1 hour is ample time to throw up issues so I doubled it. I think we can safely say the RAM is fine. Ran it with my PSU and GPU inside though, not sure if this matters.

H3llR4iser wrote: »

How did the 2080ti + PSU experiment go?

Have them in his machine since yesterday and his ones are in mine. I've had no issues and he's had one game crash. Will leave them in for a few more days and see.

.G.

He's now had 2 BSOD's with two different error codes. Only piece of hardware left at this stage is the CPU and I've had no responses to my thread on their AMDhelp reddit by anyone from AMD.

Cuddlesworth

.G. wrote: »

He's now had 2 BSOD's with two different error codes. Only piece of hardware left at this stage is the CPU and I've had no responses to my thread on their AMDhelp reddit by anyone from AMD.

Log a RMA. Put down all the troubleshooting steps you have done, isolating the cpu.

.G.

I did, they want me to send it back for testing. I don't see the point really cos I know they won't find a problem cos I can't find one myself when using aida64 and the like. If it was Amazon I'd get a no quibble replacement but this is OCUK.

Ubisoft got back to us and said the think its a Microsoft issue so we got onto MS live chat and a tech dude remotely had a poke around to analyse some recent dump files. He found evidence of a power issue and also a anticheat program that Ubisoft uses causing BSODs too. He's elevating it for another dude to check that one but recommended a new PSU for the power issue in the meantime. I've ordered one, a 650w instead of the 550w we have just to rule it out. I didn't leave my own in long as we got 2 BSODS while it was in but it seems there are multiple reasons for the issues so that may have steered me the wrong direction on the PSU.

I told OCUK I'd await the outcome of this malarkey before I bite the bullet.

AlmightyCushion

.G. wrote: »

I did, they want me to send it back for testing. I don't see the point really cos I know they won't find a problem cos I can't find one myself when using aida64 and the like. If it was Amazon I'd get a no quibble replacement but this is OCUK.

Ubisoft got back to us and said the think its a Microsoft issue so we got onto MS live chat and a tech dude remotely had a poke around to analyse some recent dump files. He found evidence of a power issue and also a anticheat program that Ubisoft uses causing BSODs too. He's elevating it for another dude to check that one but recommended a new PSU for the power issue in the meantime. I've ordered one, a 650w instead of the 550w we have just to rule it out. I didn't leave my own in long as we got 2 BSODS while it was in but it seems there are multiple reasons for the issues so that may have steered me the wrong direction on the PSU.

I told OCUK I'd await the outcome of this malarkey before I bite the bullet.

I honestly don't know how you have any of your sanity left at this stage. I'd have been driven bat shít insane ages ago if I went through all that.