Should I test this new ram kit for errors?

[Liberty610]

 

Hey everyone,

 

I just got the new Threadripper 1950x CPU.

 

The Gigabyte Aorus x399 Gaming 7 motherboard is the board I have for it, and it can be found here:

https://www.newegg.com/Product/Product.aspx?Item=N82E16813145030&cm_re=gigabye_x399-_-13-145-030-_-Product

 

The ram I have is G. Skill Trident Z 32gig whcih can be seen here:

https://www.newegg.com/Product/Product.aspx?Item=N82E16813145030

 

I also have the NZXT Kraken x62 cooler. My OS is on a M.2 Samsung 950pro SSD, and I have 5 other standard HDDs at random sizes for storage. All 7200rpm. I have the latest BIOS (f2k) installed on my board, as well as all the latest chip set drivers, ect and I am running Windows 10 Pro with the latest creative pack installed.

 

Ok so, I run a small production studio where I do a lot of music/audio work and video editing with Vegas Pro 14. Last night while rendering a video, I got a BSOD on my first pass of it. Stop Error code was: SYSTEM_THREAD_EXCEPTION_NOT_HANDLED

 

After the BSOD turned off, the computer powered down and the motherboard beep code was 3 beeps. I can't recall if it was 3 long or 3 short beeps. But they where 3 constant beeps. It then shut itself off, and turned back on, and gave me 3 more beeps. It turned off, came back on, and then 1 single beep/posted just fine.

Now, the ONLY driver that I didn't update before I had this crash was my GTX 1080 8Gb video card from MSI, which I did update right away after I got booted back up into Windows.

 

While rendering, I was watching my temps with HW Monitor and NZXT's Cam software that works with their AIO coolers, and the CPU never got past 61. I believe, maybe even a little lower was the average. I was monitoring all my temps, and everything seemed solid, so I don't believe this is an over heating issue.

 

Now before I go any further, I think I need to make mention of this, in case it was a possible reason this crash occurred. I don’t know if this setting would make JUST Vegas crash, or the entire PC. In Vegas Pro 14 settings under the ‘Video’ tab, you can set a certain amount of ram for 'dynamic ran preview'. I am assuming the higher the number this is, the faster the render goes. I have a total of 32 Gigabytes installed (4 x 8). That ram setting was up pretty high. I think it was giving Vegas something between 28gb and 29gb. When rendering starts, Vegas of course starts making use of the allowed ram assigned to it, so my ram usage shot way up. Is it possible that I fed Vegas too much ram and it caused this crash? I am not sure if it could cause a whole system crash or just Vegas to crash.

 

After the crash occurred and I rebooted, I turned the Dynamic Ram allowance in Vegas down to 16gb. It rendered the entire file just fine. It was a tad slower then the first render attempt, which is where I get the idea that less ram allowance in that Dynamic Ram setting means it goes slower. When I had the ram setting at 28gb, it estimated a 45 minuet time for the render to complete. The 16gb setting took more like 53 minuets.

 

After the render completed, I then took the video file it created, and processed it with Handbrake without any issues, and the PC was on the rest of the night into today and was working fine. Mostly in idle state, but it was running.

 

In my bios for ram, I currently have XMP enabled to get 3200mhz on these ram dimms. Gigabyte tech said to turn off XMP if another crash occurs, as the XMP overclock profile maybe causing stability issues. I have not turned that off yet, because I’m not sure what I should do yet if I need to test this ram kit.

 

Should I run Memtest86 on this new ram? And if so, do I run it with, or without XMP on? Do I test all 4 sticks at the same time in their current slots? Or do I test them one at a time? And if I test them one at a time, do I take the other 3 sticks out while I test the one stick? And do I test each stick in the same slot unless I find errors? And how long should I run these tests for? I made the USB boot for Mem86, just not sure what the best way to go about this is.

 

Thanks in advance for any replies.

[paddy-stone]

OK, I'll try to explain this simply, as I am usually long-winded, lol.

 

I wouldn't bother with memtest unless you keep getting errors, not just in one program mind as that could more likely be a software problem. So if possible run a stress test, or just re-code some video etc with handbrake etc... that should give you a baseline of whether there are actual problems with the ram or not.

 

If you do decide to do a memtest anyway, you can run all sticks if you like, then if you get errors, can start testing stick by stick on the same slot. And lastly, if you get no errors on the stick by stick test, but still had errors on the all stick test, run any stick, in slot by slot.

That's how I always did it, to get the best time efficient tests.

[iamdarkyoshi]

It never hurts to check your ram. I've gotten bad ram TWICE after RMA'ing the original set!

[Liberty610]

39 minutes ago, paddy-stone said:

OK, I'll try to explain this simply, as I am usually long-winded, lol.

 

I wouldn't bother with memtest unless you keep getting errors, not just in one program mind as that could more likely be a software problem. So if possible run a stress test, or just re-code some video etc with handbrake etc... that should give you a baseline of whether there are actual problems with the ram or not.

 

If you do decide to do a memtest anyway, you can run all sticks if you like, then if you get errors, can start testing stick by stick on the same slot. And lastly, if you get no errors on the stick by stick test, but still had errors on the all stick test, run any stick, in slot by slot.

That's how I always did it, to get the best time efficient tests.

Thank you so much for the reply! What about the XMP settings? Should I leave them on or off? And do you think that the Sony Vegas ram allowance was an issue?

[Liberty610]

I also found a couple articles that said the error was driver related, and usually related to the video card driver, which, after I got the crash, I updated my video card driver right away. The next render I tried went fine. Still haven't tried another render yet, but plan top later today.

 

https://www.techjunkie.com/system_thread_exception_not_handled-errors/
 

[paddy-stone]

3 hours ago, Liberty610 said:

I also find a couple articles that said the error voice was driver related, and usually teenaged to the video card driver, which, after I got the crash, I updated my video card driver right away. The next render I tried went fine. Still haven't tried another render yet, but plan top later today.

 

https://www.techjunkie.com/system_thread_exception_not_handled-errors/
 

Is any of that autocorrected? As that doesn't make any sense to me, lol.

 

I've found in my own experiences that most BSOD are due to driver issues. And with such a new CPU and board there could be more of these to come. So if I were you I wouldn't make any drastic changes trying to fix something that might not be broken in the first place. Do more testing without changing anything, as after you've changed something you don't know if another BSOD is indicating a problem with the original, or because you changed something.

Just test whatever you can IMO, and keep a look out for others reporting similar problems. I don't use Vegas, so can't help you there... maybe someone else can help you with that, hopefully.

[Liberty610]

I do apologize, that was heavily auto corrected and I didn't have a chance to proofread it before I posted it. Sorry about that. 

 

It should have said that I have been doing some reading up on this particular error code, and from what I saw it is mostly graphics driver related. And as I said before, right after the crash, I updated my GTX 1080 driver to the latest and I was able to get through the render without any issues. And that driver version was a very old version because Windows installed it automatically through its updates. It was version 10 something and I think I'm up to version 13+ now.

 

I called G.Skill tech support, and they said that if I wanted to test the ram I should do it one stick at a time in the first main slot and only go through a couple of passes for each stick. If I don't see any errors then I should put them all back in and go about me business. I'm also gong to leave XMP on for now as well, and do some more rendering to see if there is any more issues. if I successfully render a video using the new RAM settings I put in Vegas, then I will up the ram setting in Vegas and go from there. If it doesn't crash after that, then I'll just hopefully be good to go.

 

The only thing that is throwing me off about this crash is the weird three beep pattern that my motherboard speaker did a couple times before it would post correctly after it recovered from the crash. If it wasn't for that, I wouldn't be investigating so hard.