Question Seemingly random BSOD's

[Tanokan]

Hey!

So I made a new build back in September, and have been getting what seems to be random BSOD's since. It varies, but it happens once or twice a day, on average. It might be a coincidence, but it does not seem to happen while gaming.

I have tried updating various drivers (realtek, video, LAN as well as every single one in device manager), windows update is up to date, BIOS is up to date, a couple of weeks ago i tried resetting windows, I've run sfc /scannow, I've run chkdsk, I've tried to run memetest86 but get blackscreen everytime (tried changing boot-options in BIOS, didn't help). Tried windows memory diagnostics instead with 10 passes, extended and cache on everything, which found nothing.
I've disabled fastboot, I've changed the system power settings.
XMP-profile was not set initially, but setting it didn't help either. I've tried moving the RAM-sticks from A2/B2 to A1/B1 and then to B2/A2.

I have not done any overclocking.

Hardware:

Motherboard: Gigabyte Aorus Elite x570
CPU: AMD Ryzen 7 3700x
GPU: AMD Radeon RX 5700 XT
SSD: Samsung SSD 970 EVO 1TB
RAM: G.Skill TridentZ F4-3600C17D-16GTZSW

Dump files: https://drive.google.com/open?id=1YKyVwygJvsA_dBkZIfgi3j9LVkPeHT5x
Memory-dump: https://drive.google.com/open?id=1_xu7cNbrr77cncze-VB4kwqVFvdu8CE3

I have tried a lot of things from various threads on various forums, and still feel no closer to a solution.

I'd be grateful if anyone here can pinpoint the error.

 

[Darkbreeze]

If you don't mind my asking, where ARE you from, as I might be able to better guide you towards a good unit if I know? Any of those units is fine, I'd definitely go for the 2018 version of the RMx over the older version though. Either the RMi or RMx are better than the RM, which has an inferior capacitor selection and an inferior fan compared to the other two.

The Earthwatts Gold is a fine unit too. Not as good as the RMx, but better by FAR than what you had.


This is what our old tier list, that was fairly accurate as compared to ALL other existing or defunct tier lists, had to say. The old HardOCP review for the FSP Raider said it was a POS too, but that review is gone now because all of the old HardOCP reviews are lost now that the entire site has been taken down.

Looking at it's companions there, the quality of it's company on that level of the tier list was extremely poor. Not the worst, as they weren't tier five, but all these units listed below were just really bad power supplies. Any time I come across one of these they go straight into the trash can.

Tier Four - Not for overclocking systems or high end gaming rigs. May not even output labeled power and fail standard ATX specifications slightly. May even use cheap components to meet a price

Aerocool
GT Series 500s 700s 1050s
Integrator 600w
Strike-X series 800w 1100w
Templarius Imperator series

E-power units
Firepower Fatal1ty 2013
FSP Raider series
InWin Glacier
LC power
LEPA MX-F1 series (Trigger-happy Overcurrent protection and very poor quality capacitors)

NZXT
Hale82N 650w
Hale82 V2 700w

Thermaltake Smart series[/b]
Zalman US Fanless 400w (HardOCP reports some poor quality issues at such a high price)

 

[Tanokan]

I'm from Denmark.

And the reason I'm asking about the RMi is because the guide you linked mentioned it as better than the RMx. So if they're basically the same I might as well take the cheaper one, but otherwise $15 isn't a lot of money.

 

[Darkbreeze]

Either of them are good. The only REAL difference is that the RMi integrated the use of the Corsair link for some internal monitoring while the RMx does not have this feature. Mostly, this is not a necessary feature as it is not present on any but a handful of power supplies across the industry.

This looks like your best choice right now.


PCPartPicker Part List

Power Supply: Corsair RMx (2018) 650 W 80+ Gold Certified Fully Modular ATX Power Supply (873.00kr @ Alternate)
Total: 873.00kr
Prices include shipping, taxes, and discounts when available
Generated by PCPartPicker 2020-02-06 18:53 CET+0100

 

[Tanokan]

I have now ordered the RMx. Hopefully it arrives tomorrow, and then I'll see if that does the trick.

Thank you all so much for your help so far.

 

[Darkbreeze]

Don't be entirely surprised if the problem isn't instantly fixed, just know that THIS is the MOST likely culprit and without eliminating it as a possiblity, doing anything else is just more like chasing ghosts than anything else.

 

[Tanokan]

Oh I won't. I have a hard time getting more than cautiously optimistic on this issue at this point.

 

[Darkbreeze]

I think it's about a 50/50 chance to be honest. But if the PSU is not the problem, I think it makes it a lot easier to figure out what it IS from that point on though. Plus, given what you had as a power supply, it was really a necessity anyhow and you can sleep better at night, well, I can anyhow, knowing that your machine isn't likely to burn up anytime soon due to the PSU.

 

[Tanokan]

Yeah, even if it doesn't help solve it I'm absolutely fine with replacing the old one.

So if there is a chance it solves it, it makes it easier to troubleshoot, and gives you some peace of mind that's just a bonus

 

[Darkbreeze]

LOL.

 

[Tanokan]

So, it's been about a week since I changed PSU.
I have since then had 2 BSOD's, one Tuesday, and one Wednesday. It does seem to have improved, but I'm a little hestitant to say that it definitely has since I also went without BSOD's fri-mon the week before.

Are we back at RAM being the most likely culprit? If so I'll just buy a new set. Currently looking at these:

https://www.gskill.com/product/165/...VDDR4-3600MHz-CL16-16-16-36-1.35V16GB-(2x8GB) (I think we were looking at these the first time around as well, but they weren't in stock anywere)

https://www.gskill.com/product/165/...VDDR4-3200MHz-CL14-14-14-34-1.35V16GB-(2x8GB)

Their specs seem decent and aren't too pricey. They should also be compatible
I have also looked at the similarly specced TridentZ's, but they're more expensive, and I'm not sure what the difference is.

 

[Darkbreeze]

As per my good friend and fellow moderator Computronix:

Memory specifications and overclocking can be very deceiving. If you're unaware, here's the formula for "True Latency":

1 / Frequency (not DDR) x Latency = True Latency (nanoseconds).

Stock 3200 @ 14 is faster than Stock 3600 @ 16:

1 / 1.600GHz x 14 = 8.75nS
1 / 1.800GHz x 16 = 8.89nS

Stable Overclock with 3733 @ 16 is faster:

3733 @ 16 is 1 / 1.867GHz x 16 = 8.57nS

So I'd go with the CL 14 3200mhz kit, which in fact is a Samsung B-die kit anyhow, meaning it is a terrific memory kit using the absolute best memory chips on the market. Regardless that it is Ripjaws and not Trident Z or Dominator Platinum, they are still superb memory modules and should actually be faster than the 3600mhz kit in terms of actual latency, which is where the rub lies anyhow.

 

[Tanokan]

A'ight, I've just ordered the 3200's, I can hopefully pick them up tomorrow or Sunday, and then we'll see how it goes.

 

[Darkbreeze]

Fingers crossed.

 

[Tanokan]

Swapped out the RAM-sticks wednesday (and also made sure XMP was enabled), and just got another BSOD. Although i think this is the first time I've gotten an ATTEMPTED_EXECUTE_OF_NOEXECUTE_MEMORY, but I'm not sure.

 

[Darkbreeze]

Did you do a hard reset of the BIOS after installing the new sticks? If not, I would do that. Power off. Unplug the cable or switch the PSU to the "0" position. Pull the CMOS battery for five minutes. Press the power button for 30 seconds continuously during that five minutes. Put the CMOS battery back in. Plug back in or switch PSU back to the "I" position. Power on. Go into BIOS and enable XMP. See how it does.

Do you have a Windows 7 or 8 driver installed for any piece of hardware or peripheral, maybe one that doesn't HAVE a Windows 10 driver available?

 

[DMAN999]

That new RAM kit is listed by G.Skill as compatible with your MB so I would clear the CMOS using the CMOS jumper shown on page 15 of your manual:
https://download.gigabyte.com/FileList/Manual/mb_manual_x570-aorus-elite-wifi_e.pdf
Using these steps:

1. Unplug the PC (PSU).
2. Then press the Power Button for at least 30 seconds.
3. Use a flat headed screwdriver to short the CLR CMOS Jumper (Touch both jumper pins with the screwdriver) for at least 15 seconds.
4. Plug the PC back in.
5. Press the Power Button to turn the PC on.
6. Repeatedly tap the DEL key
7. A screen should come up telling you to press the F1 key, do that.
8. Set the Optimized Default settings and Press OK (this is on the EXIT tab)
9. Save and exit the BIOS.
10. Tap the DEL key again to enter the BIOS and change any settings you want changed then save and exit again.

Then try a clean re-install of Windows 10 to eliminate a software (OS) issue.

 

[Tanokan]

Tried Darkbreezes approach, BSOD'ed within 15 minutes, and the tried DMAN999s approach including clean install.
So far I've let Windows Update do its thing, installed radeon software 20.1.3, Google Chrome, Discord and League of Legends. Haven't touched Device Manager or the drivers on the motherboards page at GIGABYTE.

Let's see how it goes.

 

[Darkbreeze]

If you recall, I recommended doing a clean install much earlier in this thread. Here in fact:

https://forums.tomshardware.com/threads/seemingly-random-bsods.3569159/post-21559809

And you said you did a clean install in September and a reset a few months later. Problems like this, I always recommend doing a clean install because then you at least know you are at square one.

 

[Tanokan]

I remember. My previous comment wasn't meant as a dig in any way. It was just a description of what happened.

Anyway.. I've so far crashed twice.

Dumps for anyone interested: https://1drv.ms/u/s!AnVqCzqrKBXsqaFtb-zI9Q81ynhelA?e=SsN9gn

 

[gardenman]

I tried to download the memory.dmp and it failed (right as it finished). It took me about an hour due to my slow connection. I'll try again later if needed, but I get the same info from a full memory.dmp as I do the normal minidumps. There are others around who can get more info from the memory.dmp but they are not around as often.

Minidump results: https://silkiestcaterpillar.htmlpasta.com/

File information:	022420-5515-01.dmp (Feb 24 2020 - 13:12:17)
Bugcheck:	IRQL_NOT_LESS_OR_EQUAL (A)
Probably caused by:	memory_corruption (Process: System)
Uptime:	0 Day(s), 21 Hour(s), 31 Min(s), and 27 Sec(s)

File information:	022220-5421-01.dmp (Feb 22 2020 - 13:10:08)
Bugcheck:	REFERENCE_BY_POINTER (18)
Probably caused by:	memory_corruption (Process: chrome.exe)
Uptime:	0 Day(s), 2 Hour(s), 42 Min(s), and 18 Sec(s)

This information can be used by others to help you. I can't help you with this. Someone else will post with more information. Please wait for additional answers. Good luck.

 

[Colif]

I may be mistaken but I think Reference by pointer is normally an internet problem. I would update the LAN drivers on motherboard as yours are from 2018 and latest on https://www.gigabyte.com/au/Motherboard/X570-AORUS-ELITE-rev-10/support#support-dl-driver-lan are 2019.

i can't tell anything useful from error 1, but as GM said, if you lucky @PC Tailor might come back or @axe0axe0 might not be too busy to look at dumps for you.

 

[Tanokan]

I'll get on it when I get home. I purposefully skipped it to limit moderations as it hasn't changed much in the past, but that was before new PSU/RAM.

 

[Tanokan]

Updated LAN-driver to the one specified on the motherboard support page, and crashed again not long after. Dump file added to the previously linked one-drive folder.

 

[gardenman]

I ran the new dump file through the debugger and got the following information: https://garretedtoaster.htmlpasta.com/

File information:	022520-5421-01.dmp (Feb 25 2020 - 11:42:35)
Bugcheck:	SYSTEM_SERVICE_EXCEPTION (3B)
Probably caused by:	memory_corruption (Process: Discord.exe)
Uptime:	0 Day(s), 22 Hour(s), 14 Min(s), and 05 Sec(s)

This information can be used by others to help you. I can't help you with this. Someone else will post with more information. Please wait for additional answers. Good luck.

 

[Darkbreeze]

I would uninstall Discord and see if you continue to have problems. If not, then it is probably because Discord, much like Skype used to, might tend to be very buggy at times. I've seen a fair number of people here lately with random BSOD problems that were using Discord. Never heard back from any of them after suggesting that might be the problem but I think that alone makes it at least moderately likely it was. Could also be that it is conflicting with something else on your system. Could also be that some resource it is using is the actual problem, but as a test it's probably worth avoiding for a day or two to check if it goes away.