• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Persistent issue with freezing; elusive diagnosis

J

John Tzimisces

Junior Member
Hello, I had this place recommended to me by a friend for a persistent issue I've had with the computer in question and my previous machine.

In my previous machine, I experienced intermittent freezing and BSODs while either gaming. At first it was associated with when the machine had been off for a while (I'd get a freeze or BSOD or two and then it'd run smoothly for pretty much however long) and then more recently, I'd get a freeze or BSOD every day while gaming, and then while not gaming, and then more than once a day.

I attributed this to a slowly failing power supply, or the memory, so I replaced the mobo, memory, PSU, and CPU. The build went smoothly and everything actually appeared to be going great after I wiped the boot drive and reinstalled windows 10. A little weirdness during the process due to using a USB 3.1 (I think?) port with the install media that was resolved by starting over with a USB 2 port.
However a couple days ago I dipped my feet back into the closest thing I have to graphics intensive gaming, running fallout 4, and the freezes returned albeit after hours of gaming. The second or third one I got in as many days actually forced me to use my USB windows install media to rebuild winload.exe (which was really scary as that hasn't ever happened before).
Freezes have not happened while not gaming with the new build.
I'm also not experiencing any weird artifacts and otherwise things are running smoothly, except for one weird thing where FO4 simply crashed when first running it, and the computer repeatedly informed me that the graphics didn't have permission to interact with the program, or something to that effect.

I'm investigating a couple free options but if you guys have opinions I'd really appreciate it. I haven't made any tweaks to BIOS yet and have no intention to overclock (I don't think the H170 mobo I have is suited to it anyway). I don't allow the computer to sleep, and I use the default balanced power plan otherwise.

I've replaced the surge protector because the old one is well over a decade old, but the issue persists.
Next, I'll be replacing the SATA cables for the SSD and HDD with cables that came with the mobo (I have been working on the assumption that the ones currently in use aren't old, since both of those drives were installed a year ago).

Mobo: MSI H170 gaming M3
Memory: Corsair Vengeance LPX 16GB DDR4-2133
PSU: EVGA 750B2 semi-modular (there are no old power cables in the case, other than a weird connector on the graphics card, which allows two 6 pin connectors to connect to the card)
CPU: Intel i5 6500 3.2 GHz
Case: Corsair vengeance C70

The parts which I retained from the previous machine:

SSD : http://www.newegg.com/Product/Product.aspx?Item=N82E16820148820 (I'm pretty sure)
HDD : http://www.newegg.com/Product/Product.aspx?Item=N82E16822236625 (I'm pretty sure, again)
Graphics: GTX 960 2GB (unsure what manufacturer, I'd have to open up the case again)
and the optical drive.

I really appreciate any and all opinions; until this point I was really proud to have been able to assemble this as my first PC build.
 
PSU: EVGA 750B2 semi-modular (there are no old power cables in the case, other than a weird connector on the graphics card, which allows two 6 pin connectors to connect to the card)
Click to expand...

You do have the pcie power cable from your PSU plugged into to the both of the 'weird connectors' on your video card, correct?
 
XavierMace said:
What are your temps like? Do you have any errors in the event log?
Click to expand...
I would have to record a log with open hardware monitor or CPU-Z, right? As far as I know nothing shows up in the event log, although I've just been looking under windows logs -> system in the event viewer.

UsandThem said:
You do have the pcie power cable from your PSU plugged into to the both of the 'weird connectors' on your video card, correct?
Click to expand...
I have two cables (the kind which are 6 pin but have an extra 2 pins in case it needs to interface with 8) from the PSU connecting to the two connectors coming off the video card.

VirtualLarry said:
Sounds like the elusive "Skylake freezing issue". Consider replacing your CPU.
Click to expand...

...a cursory googling of "skylake freezing issue" seems to indicate I ought to just flash the BIOS?
 
John Tzimisces said:
I would have to record a log with open hardware monitor or CPU-Z, right? As far as I know nothing shows up in the event log, although I've just been looking under windows logs -> system in the event viewer.


I have two cables (the kind which are 6 pin but have an extra 2 pins in case it needs to interface with 8) from the PSU connecting to the two connectors coming off the video card.



...a cursory googling of "skylake freezing issue" seems to indicate I ought to just flash the BIOS?
Click to expand...

You might have already mentioned it and I missed it, but have you checked which BIOS version you are running, and checked at MSI's website to see if there is any updated BIOS that fixes system stability?

Also, since the crashes happen during gaming, download the Intel Extreme Tuning Utility and run some stress tests on it (CPU, memory, GPU) and watch what your temperatures do (this program shows real-time temps). I think running each one for 15 minutes would show any issues if it was heat related.
 
UsandThem said:
You might have already mentioned it and I missed it, but have you checked which BIOS version you are running, and checked at MSI's website to see if there is any updated BIOS that fixes system stability?

Also, since the crashes happen during gaming, download the Intel Extreme Tuning Utility and run some stress tests on it (CPU, memory, GPU) and watch what your temperatures do (this program shows real-time temps). I think running each one for 15 minutes would show any issues if it was heat related.
Click to expand...

Thanks for lead, I downloaded the utility. As for BIOS, my version comes from 9/6/2015 (AMI C.10?). MSI's website has a few versions newer than that.

edit: under stress test, I don't appear to have an option for GPU?
 
Last edited:
John Tzimisces said:
Thanks for lead, I downloaded the utility. As for BIOS, my version comes from 9/6/2015 (AMI C.10?). MSI's website has a few versions newer than that.
Click to expand...

Must be version C1. They are now up to C5.

A lot of fixes over the revisions for CPU microcode, PCIe compatibility, VGA compatibility, memory, and a lot of stuff.

Go ahead and update your BIOS to C5.

edit: under stress test, I don't appear to have an option for GPU?
Click to expand...

It's probably not there since you are using a graphic card instead of Intel's integrated graphics. No biggie on that test. There are plenty utilities that can stress your video card.
 
Last edited:
A GTX960 has two 6 pin power connectors?

I would expect a single 8 pin for the models that creep up past 150W?
 
Last edited:
VirtualLarry said:
A 6-pin and an 8-pin? On a GTX960? Crazy...
Click to expand...

That's what I told my son. 160w for a 960? For that power draw, I convinced him to wait until the new Pascal cards hit, and all the 'must have newest card' early adapters flood the market with their used GTX 970s.

That's the only reason I'm excited about the Pascal launch. 🙂
 
I was a little perplexed by it (the two 6 pin thing) myself. I haven't really fiddled with it as it looks like its some kind of adapter.

In any case I'll be flashing the BIOS in a little bit, and then running a stress test using that utility afterwards.
 
John Tzimisces said:
I was a little perplexed by it (the two 6 pin thing) myself. I haven't really fiddled with it as it looks like its some kind of adapter.

In any case I'll be flashing the BIOS in a little bit, and then running a stress test using that utility afterwards.
Click to expand...



I'd start with the Intel utility I mentioned earlier so you can see your CPU's temperature range.

You can use this to stress your graphic card. Depending on the airflow of your case, you might want to leave the side panel off because it will crank up the heat.

https://unigine.com/products/benchmarks/valley/
 
Well, turns out I had time to flash before I saw that. I suppose I should move on to the CPU stress test (I did note that the resting temperature was around 30C not that that really matters).
 
John Tzimisces said:
Well, turns out I had time to flash before I saw that. I suppose I should move on to the CPU stress test (I did note that the resting temperature was around 30C not that that really matters).
Click to expand...

Yeah, you really just want to see what it climbs to after 15 minutes or so. As long as doesn't get too high, you could rule out heat causing issues for your CPU.

It's all just really a process of elimination, and always starting with the easy stuff first to rule those out.
 
The 15 minute CPU stress test peaked at 70 and hovered around 67-69C, the memory test had it going between 62 and 72C. That seems reasonable, at least for a stock cooler.
 
John Tzimisces said:
The 15 minute CPU stress test peaked at 70 and hovered around 67-69C, the memory test had it going between 62 and 72C. That seems reasonable, at least for a stock cooler.
Click to expand...

Yeah, unfortunately that's the norm for stock Intel coolers.

You can run the Valley benchmark for a bit (30 minutes or so).

If all goes fine with it, go ahead and use your computer as you normally would and see if you encounter another BSOD or other odd behaviour. If you do, report back here with what you were doing when it happened, and we will go on from there.
 
Valley benchmark on Extreme HD had the GPU running at like 68C for the full half hour. At this point I'm feeling pretty good! I'll report back if anything catches my eye (or if nothing catches my eye, after X amount of time)
 
John Tzimisces said:
Valley benchmark on Extreme HD had the GPU running at like 68C for the full half hour. At this point I'm feeling pretty good! I'll report back if anything catches my eye (or if nothing catches my eye, after X amount of time)
Click to expand...

Please do report back even if you are good to go.

Threads like these help others out when they are doing keyword searches for similar problems they are experiencing.

Good luck!
 
Hmm, no dice. Just had to hard shut down again.

I can try again, but this time keep a monitor recording or something, right? (although will it save if the computer just locks up like it does?)
 
John Tzimisces said:
Hmm, no dice. Just had to hard shut down again.

I can try again, but this time keep a monitor recording or something, right? (although will it save if the computer just locks up like it does?)
Click to expand...

Did it crash while gaming?

Ok. So here is where we are at. Let me know if my summary is incorrect:

-You have replaced all of your hardware at some point while trying to figure out the problem.

-You have done a least one clean re-install of Windows 10.

-You have gone to MSI's website and installed the drivers they have there for motherboard (chipset, inf, USB, SATA, sound, etc).

-You have downloaded an installed the newest driver directly from Nvidia?

-----------------------------------------------------------------------------------------

Please right click on your start button, select control panel, select system and security, select security and maintenance, click maintenance, and finally select view reliability history.

What shows up in that history? Do you have a lot of red x's (mainly looking for the reds (application failures, Windows failures, warnings). You don't have to list them all, but is there one that appears pretty consistently?
 
What was it doing when it crashed? Did it create a minidump? Whocrashed can be helpful in finding out what is causing these minidump files.

Another thing to keep your eye on is voltages. CPUID HWMonitor can show you what your voltages are like with loads (particularly cpu and 12v).
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top