Random crashes on Asus x570+Trident Z Neo 16GTZNC

[DieVirulenz]

TL/DR: New Ryzen System reproducibly crashes a few applications(mainly Blender, most run perfectly fine) , RAM timing issue or RMA case?

Hi,

for about a month I got my new Ryzen 3 (Zen 2) system. For the most part is runs buttery smooth and as expected. Specs are:

Ryzen 3900x
Asus x570 Strix-E Gaming, 1.0.0.3 ABBA updated
G.Skill Trident Z Neo, F4-3600C16D-16GTZNC, Hynix CJR
Nvidia GTX 1080
bequiet 680W PSU

BIOS-settings are on default(updated to ABBA 1.0.0.3) except I activated the DOCP profile for my RAM (which is on the QVL of the MB, btw.). Chipset drivers are installed both first from AMD directly then from ASUS again, as is stated in the description of them on the ASUS page).

Played some games (Mordhau, Greedfall, Battlefront 2, several others) everything works fine. Rendering with FFMPEG works fine too.

So a few days ago I installed Battlefield 5 to see how it performs (hadn't played on my old rig in month) and it crashes to desktop, sometimes after a minute, sometimes after one or two matches. As this is Battlefield which has been prone for crashes like this I though lets try some more things, and a few games and tools really show similar behavior after trying quite a few from my library (for example "Exanima" starts on DOCP V1 but just freezes after a few seconds, "Between the Stars" shows a similar behavior).

What really shocked me was when I installed Blender and fired it up and it just crashes after a few seconds (maybe 10-30).

I began fiddling with my RAM timings, tried BF and Blender with some different settings: all on AUTO(so 2133HZ), manually set DOCP values and suggested values from Ryzen RAM Timing calculator both "Safe" and "Fast. The behavior stayed the same, some settings felt a bit more stable at first but in the end it stays the same.

Note: I never had a BSOD so far, and I monitored my temperatures for processor and GPU, which are both well inside specifications, in case that helpes pinpointing.

Now I am at a loss, I ran a memtest from USB stick over night with DOCP activated and this did not return any errors, I really am not sure what the culprit might be now. Does anyone maybe have a similar or even the same combination of RAM and MB and might sahre some comparison on behavior?

Might it be a BIOS issue, or rather some DRAM timings? Havent really overclocked in years, so no real experience regarding this.

Maybe someone here know about similar issues or got any ideas, let me know if I can supply any more information regardings this to track this down.

Thanks in advance!!!

 

[daoson5]

Hi, can you borrow some one have a PSU (750 or 850 w )

 

[DieVirulenz]

I dont think so, do you suspect 680w might not be enough? I might just get a bigger one and try that, but I figured it should be enough.

Anyways thanks, I'll keep it in mind!

 

[oobymach]

If you're running the newest driver consider downgrading to the latest studio drivers, they're more stable and could resolve the issue.

Also battlefield especially is designed to take full use of your cpu, it may just be overheating and throttling.

 

[DieVirulenz]

If you're running the newest driver consider downgrading to the latest studio drivers, they're more stable and could resolve the issue.

I am not certain what you mean with "studio drivers"?

Also battlefield especially is designed to take full use of your cpu, it may just be overheating and throttling.

I will recheck the temps once more, just to be sure. Also shouldn't BF2 be utilizing the CPU in the same way then?

Anyways thanks, will check temps again.

 

[oobymach]

Download The Latest Official GeForce Drivers

Download the latest official GeForce drivers to enhance your PC gaming experience and run apps faster.

www.geforce.com

In the results list after you select your hardware will be Nvidia Studio Driver, try the latest one to see if there's a difference.

 

[DieVirulenz]

Ahh okay, so thats some kind of LTS driver I assume?

Will give it a try later and report back, thx!

 

[Agent_D]

What other items do you have in your system? The 680w supply may be the culprit if you've got several SSDs and/or HDDs or other components using power. You said temperatures are within spec, what would those be? Have you monitored your CPU clocks and the voltage/current with AMD Ryzen Master? Have you tried installing memory in different slot configurations? 1/3 or 2/4?

 

[DieVirulenz]

So the driver didn't make a difference.

What other items do you have in your system? The 680w supply may be the culprit if you've got several SSDs and/or HDDs or other components using power. You said temperatures are within spec, what would those be? Have you monitored your CPU clocks and the voltage/current with AMD Ryzen Master? Have you tried installing memory in different slot configurations? 1/3 or 2/4?

I got 1NVME for my sys, two more SSDs and two HDDs, chassis and CPU: 4. But usually when I game they should just idle. Should...also I don't think that for example when I start Blender that much watts will be needed, still it crashes.

Temperatures (tops)after two hours Battlefront 2: 76°GPU, 66°CPU

Havent monitored voltages, what would I look out for? Spikes of some kind?

I currently have the DIMMs in 2/4, which is suggested by the mainboard. I will go ahead and try to switch it to 1/3 and see if that helps. I will also try to get some other DDR4s and switch them

 

[TheLostSwede]

In my experience, all the game crashes I've had, have been related to memory timings.
Try the DRAM calculator and safe settings to see if that helps.

DRAM Calculator for Ryzen v1.7.3 Download

DRAM Calculator for Ryzen helps with overclocking your memory on the AMD Ryzen platform. It suggests stable memory timing sets optimized for your m

www.techpowerup.com

 

[DieVirulenz]

I tried them already, it felt a bit different, but in the end all the crashes still occurred.

I'm going to try to switch to slower timings later, like: 16-20-20-40 and give it a shot.

But I also feel that it has to be related to RAM in some way. As all of these components are rather new I am just not sure if its just a glitch in some software (BISO, Windows, anything really) or if its a RAM malfunction on HW level.

Well off to some more testing and thanks so far for all the help, really appreciated!

 

[eidairaman1]

Are you getting BSOds at all or just restarts?

If it auto restarts you need to disable auto restart upon crash for memory, bump the ram voltage up a little.

If no dice RMA them.

 

[DieVirulenz]

No, no bluescreens at all. Only a few applications crash reproducibly.

The system doesn't freeze either or anything.

 

[eidairaman1]

Have you checked the event viewer?

When the apps crash do you get a error on screen?

 

[DieVirulenz]

Yes I have, there are errors when something crashed. I would have to check those though when I get back home.

 

[thesmokingman]

Did you test each stick individually yet?

 

[eidairaman1]

Yes I have, there are errors when something crashed. I would have to check those though when I get back home.

Write those errors down for each app that crashes

 

[Zach_01]

...and could not hurt to monitor the system with HWiNFO64 (sensors only) for anything odd. Gives a tone of temps, voltages, wattage of all components.
Assume you're running Win10? Have you checked the EventViewer? Right click the Win10 logo down left and you can find it.

 

[TheMadDutchDude]

Your PSU is definitely more than enough for that system... don’t pay attention to them.

In my experience, random crashes are typically voltage related, IMC to be more specific. Mine is finally dialed in to perfection, but it took a lot of tweaking.

 

[DieVirulenz]

TLDR: RAM was faulty!! In just a few applications, especially OpenGL it crashed leading to the Graphicscards driver. But it was just the RAM in the end.

Soo I had to step back a day after trying out all kinds of solutions, everything to no avail, and getting way too frustrated. I ordered myself a break...and a new kit of RAM.

Well long story short, after I switched the RAM everything runs buttery smooth and so far no crashes in all applications which crashed continously beforehand.

Therefor I wish to thank you all again for your support and ideas, it helped a lot and I really appreciated it, just wanted to let everyone know what the culprit was in the end.

 

[cucker tarlson]

TLDR: RAM was faulty!! In just a few applications, especially OpenGL it crashed leading to the Graphicscards driver. But it was just the RAM in the end.

Soo I had to step back a day after trying out all kinds of solutions, everything to no avail, and getting way too frustrated. I ordered myself a break...and a new kit of RAM.

Well long story short, after I switched the RAM everything runs buttery smooth and so far no crashes in all applications which crashed continously beforehand.

Therefor I wish to thank you all again for your support and ideas, it helped a lot and I really appreciated it, just wanted to let everyone know what the culprit was in the end.

same kit ?

 

[DieVirulenz]

same kit ?

Good point and no its a Corsair Vengeance Pro (CMW16GX4M2Z3600C18) now, also from the QvL for my board.

 

[cucker tarlson]

Good point and no its a Corsair Vengeance Pro (CMW16GX4M2Z3600C18) now, also from the QvL for my board.

lol,and people are saying corsair is the worst for ryzen
IMO this wasn't necessarily the ram was faulty,just the IMC on Ryzen is friggin picky when above 3200.

 

[DieVirulenz]

Well, hopefully it will behave now, and I do agree, but if it runs now I'm happy. Thx!

 

[Calmmo]

Whenever you build a system try make some time for a bit of ram/cpu/gpu stability torture to ensure they run stable at stock. Like an hour or two of memtest and another couple hours of CPU / gpu benches. Just to make sure no errors or throttling occur vrm getting really hot etc.
I would still run tests on the old sticks of ram to see if and which one is causing the problems.