Question Is this a Ram problem?

[cireem]

I built my new system about 3 months ago, enabled xmp about 2 months ago. Never had a problem until today.

Twice both of my monitors went black while surfing the web. The first time after going black they both powered off then came back on. The second time I heard 3 rapid beeps and then they did the same thing.

I have a x570 aorus elite, checking online 3 beeps seems to mean a memory problem.

First I tried reseating the ram as I had recently moved my tower (gently.)

I ran memtest86 when I enabled xmp and had no problems.

Looking for a quick diagnosis I ran memtest64 and it can trigger the monitors blacking out within 1-10 minutes. The interesting thing is they dont power off when memtest is running. They stay black with the windows bar at the bottom showing(unlike the complete black and power cycle I got when browsing the web.) I have thr background set to change every minute on my computer so once a minute it flashes the background then goes back to black with the menu bar showing.

I ran testmem5 with anta777 extreme config and got 0 errors.

I disabled xmp, ran testmem5 again, no errors. Ran memtest64 and got the black screen instantly.

I cant see if memtest 64 is saying I have any errors because of the black screen.

Any help would be most appreciated.

Edit: Running memtest86 with xmp still off now.

 

[Lutfij]

Considering that this is a thread of troubleshooting nature, it's customary to include your full system's specs. Include the version of BIOS you're on and the rams you're using. If X.M.P is causing the issue, rule that out by manually inputting the ram's advertised frequency, timings and voltage.

 

[cireem]

Considering that this is a thread of troubleshooting nature, it's customary to include your full system's specs. Include the version of BIOS you're on and the rams you're using. If X.M.P is causing the issue, rule that out by manually inputting the ram's advertised frequency, timings and voltage.

Oh right! Sure thing.

Ryzen 3800x
X570 aorus elite
Crucial ballistix 3600 ddr4 2x8g (same package)
3070 ventus 3x oc
Evga 750w g2
Windows 10

I can get more precise specs like the bios and revision when memtest86 is done.

I had only used xmp to get the advertised frequency. I reverted back to the stock settings when I last ran memtest64 and it's still on them while running memtest86.

 

[cireem]

Ran memtest86 with 4 passes twice. Took about 12hours. No errors. I can still replicate the crash by running memtest64.
Bios version is F11
Ram is BL2K8G36C16U4B

I'd really appreciate it if anyone has any input.

 

[Third-Eye]

Ran memtest86 with 4 passes twice. Took about 12hours. No errors. I can still replicate the crash by running memtest64.
Bios version is F11
Ram is BL2K8G36C16U4B

I'd really appreciate it if anyone has any input.

Are you certain your ram is actually running with XMP enabled for 3600 CL16? Have you confirmed the speed with a program like CPU-Z or HWInfo when you enabled XMP? No errors for 12 hours when you kept having issues with ram could mean the motherboard is setting the ram back to default 2133.

Also, Bios F21 has improved memory compatibility. you dont really need to upgrade past F21 since you only need the newest bios if you plan to run Ryzen 5000 right now.

 

[cireem]

Are you certain your ram is actually running with XMP enabled for 3600 CL16? Have you confirmed the speed with a program like CPU-Z or HWInfo when you enabled XMP? No errors for 12 hours when you kept having issues with ram could mean the motherboard is setting the ram back to default 2133.

Also, Bios F21 has improved memory compatibility. you dont really need to upgrade past F21 since you only need the newest bios if you plan to run Ryzen 5000 right now.

2 months ago I saw it was running at 2133 in hwinfo so I switched it to its advertised settings via xmp. I ran memtest64 and 86 at that point. It was stable for both with 0 errors. Now two months in I'm getting black screens.

I reverted it to 2133 and disabled xmp to see if that would fix it. I still get black screens when using normally and can trigger the black screen by running memtest64. Meanwhile running memtest86 and testmem5 at 2133 are still giving me no errors.

Should I try upgrading the bios?

Is it possible it's gpu related since the problem seems to be with the output?

 

[Third-Eye]

2 months ago I saw it was running at 2133 in hwinfo so I switched it to its advertised settings via xmp. I ran memtest64 and 86 at that point. It was stable for both with 0 errors. Now two months in I'm getting black screens.

I reverted it to 2133 and disabled xmp to see if that would fix it. I still get black screens when using normally and can trigger the black screen by running memtest64. Meanwhile running memtest86 and testmem5 at 2133 are still giving me no errors.

Should I try upgrading the bios?

Is it possible it's gpu related since the problem seems to be with the output?

You were having 3 beeps, which is related to a memory issue. It's very possible that when you are running stable, your ram is at 2133 and XMP failed, so I would update the bios to F21 and try XMP again. Personally I found memtest64 to be less reliable than memtest86 and I really only used testmem5 when I was doing memory timing tweaks.

 

[cireem]

You were having 3 beeps, which is related to a memory issue. It's very possible that when you are running stable, your ram is at 2133 and XMP failed, so I would update the bios to F21 and try XMP again. Personally I found memtest64 to be less reliable than memtest86 and I really only used testmem5 when I was doing memory timing tweaks.

I only heard the three beeps once. All the other times it's been the same but no beeps.
It's just weird that memtest64 can trigger the crash within minutes even at 2133 and none of the others even get an error. I can't find a log for memtest64 either so who knows what's going on.

I'm going to try updating the bios after occt finishes running. It's got 2 minutes left and hasnt found any errors either.

I'll post an update after. Thanks for your help so far!

 

[cireem]

You were having 3 beeps, which is related to a memory issue. It's very possible that when you are running stable, your ram is at 2133 and XMP failed, so I would update the bios to F21 and try XMP again. Personally I found memtest64 to be less reliable than memtest86 and I really only used testmem5 when I was doing memory timing tweaks.

Alright so I tried a couple things and it's still happening.
I uninstalled all my gpu drivers using ddu and reinstalled the newest. Tested to see if mem64 could trigger the crash after. It still does.
I updated the bios to f21. Tested again with mem64 and it still triggers a crash.
This was still with the ram at 2666 and xmp disabled.
I really have no idea what to do now.

 

[Third-Eye]

Alright so I tried a couple things and it's still happening.
I uninstalled all my gpu drivers using ddu and reinstalled the newest. Tested to see if mem64 could trigger the crash after. It still does.
I updated the bios to f21. Tested again with mem64 and it still triggers a crash.
This was still with the ram at 2666 and xmp disabled.
I really have no idea what to do now.

Manually set your ram to 3600 CL16 1.35v and the rest of the timings advertised. Leave all other timings on automatic. Try running memtest86 for at least one full default 4 test pass. If it passes, run testmem5 to check it.

edit - Basically you need to make sure that your ram is actually running 3600 CL16 in Windows with CPU-Z or HWinfo so you can make sure the memory testers are running the tests at that speed. Don't bother testing memetest64 if it keeps crashing as it possibly means something is wrong with the tester. If you get a full pass on your ram after using memtest86 and testmem5 at 3600 CL16, you likely have no problem with your ram and you have an issue with some other part(s).

 

[cireem]

Manually set your ram to 3600 CL16 1.35v and the rest of the timings advertised. Leave all other timings on automatic. Try running memtest86 for at least one full default 4 test pass. If it passes, run testmem5 to check it.

edit - Basically you need to make sure that your ram is actually running 3600 CL16 in Windows with CPU-Z or HWinfo so you can make sure the memory testers are running the tests at that speed. Don't bother testing memetest64 if it keeps crashing as it possibly means something is wrong with the tester. If you get a full pass on your ram after using memtest86 and testmem5 at 3600 CL16, you likely have no problem with your ram and you have an issue with some other part(s).

Thanks, Third-Eye, I think I may have tried that iteration but I can't remember. I'll give that a shot.
I've been using memtest64 because it was a reliable way to reproduce the crash.
I've now tried both sticks separately in the a2 slot and while one of them still crashes when I run memtest64 the other doesn't. I ran the one that doesn't crash by itself for the last few hours and I can't even reproduce the crash using memtest64.

I'm now leaning to believing that it might just be a faulty stick.
Your thoughts?