Random freezes with seemingly random intervals.

Rebornneo

Honorable
Nov 17, 2013
5
0
10,510
Hey everyone,

I've been looking/lurking around the internet for a few days now but have failed to find a solution for my problem.

What's happening exactly is that after using my pc anywhere from a few minutes (3) to a few hours (5), the screens (I have 2) freeze randomly leaving me with no control at all. This forces me to reboot, leaving no error in the system log. I also turned on Hyper-V, a common suggestion, which made no difference.
I should point out that this has only been happening after the 8.1 update, but also recently I put in a new processor (FX9590) in my mobo (Crossair V Formula-Z) which has been thoroughly tested with my previous FX8350 without trouble. I can't imagine this to be the troublemaker, especially after running Memtest and having it give me a 100% OK after three passes.
There might be the graphics card, but I currently have no alternative to test it with. I have run furmark/games on settings that make the card work 100% for tens of minutes without it freezing the PC.
What does seem to cause the freezing more often is having a video open on two screens/a video and a game on separate screens.

I'm all out of ideas and frustrated, and open for any suggestions.
(That reminds me, I'm running a Samsung 120GB 840 with a 500GB Seagate Barracuda next to it. I have a spare HDD I could test, but haven't had the time to do that yet. If you guys think it's worth it I'll do that.)

Regards,
Rebornneo
 
Solution
Hyper-V is Microsoft's answer to VirtualBox and VMware Player/Workstation. It has nothing to do with your freezing problem.

Download HWMonitor and check your temps. Once you establish a baseline at idle, download Prime95 and run the torture test. Should run all of the CPU cores at 100%. Let it go for 20-30 minutes and see how hot the CPU gets, and see if the system freezes again.
Hyper-V is Microsoft's answer to VirtualBox and VMware Player/Workstation. It has nothing to do with your freezing problem.

Download HWMonitor and check your temps. Once you establish a baseline at idle, download Prime95 and run the torture test. Should run all of the CPU cores at 100%. Let it go for 20-30 minutes and see how hot the CPU gets, and see if the system freezes again.
 
Solution

Rebornneo

Honorable
Nov 17, 2013
5
0
10,510


[Edit:] And thank you for your quick reply!

Well, I downloaded HWMonitor and noticed that my cpu was running a little hotter than I'd like at 41C.

After seeing everything remain constant for a few minutes I started Prime95 Blended 8-threaded as it was default and seemed right. Immediately I saw the CPU temp spiking up to 65 before - you guessed it - a freeze.
Of course the first thing I checked was to see if the cooler was attached properly. Judging by the amount/spread of the paste and checking the contact between cooler and cpu, it was.
(Obligatory: I should've known better than to use a stock cooler in my enthusiasm / blah)

Fortunately I have easy access to other coolers to test, I will update tomorrow after testing. (In the mean time, would you have any suggestion for a good replacement as a permanent purchase? I don't mind spending a little money on a good one, keeping in mind that I'll have an R9 290x with a water block soon.)

If you have any other comments/suggestions, I appreciate them. Thank you so far and in advance.

Regards,
Rebornneo
 
If you have any replacement paste available, you could always remove the stock stuff and reapply...

As for a decent air cooler, I can't really say. I'm a bit out of the loop on air cooling, as I mainly use closed-loop liquid coolers like the Corsair H80 now.
 

BigAlanM

Honorable
Nov 18, 2013
1
0
10,510
I've been having the same problem with my HP Pavilion AMD Quad 64bit desktop.
Seemingly randomly the cursor would freeze on the screen and neither my mouse nor keyboard would work. Usually this occurred every 15 minutes or so.

I should mention that while the input devices were inoperable I had three widgets on the screen that continued to update. ie UPS status, CPU Usage, and Clock/calendar, so the computer was working normally, it was just the mouse and keyboard that were dead.

This all seemed to start just after Windows 7 performed it's Tuesday night update.
I was sure that it was a software problem or a hard drive problem so spent two days chasing those possibilities.

It finally dawned on me that the freeze always occurred when I was moving the mouse a certain way in a certain area of my, cluttered, desk.

I then remembered that a couple of years ago I had similar problems. Not a freeze, but Windows would randomly beep and show an information window saying that a driver had been installed for a new USB device.

Long story short. My cords of my corded mice are wearing out and intermittently, when pushed/stressed in a certain direction, opening/shorting. This, somehow, kills the USB ports that the mouse and keyboard are plugged in to.

I replaced my mouse with a spare and haven't seen the freeze problem in 36 hours.

I WILL be switching to a cordless mouse in the near future.

 

Rebornneo

Honorable
Nov 17, 2013
5
0
10,510
[Last post: So, my update for the day... sad, but fixable news.

I put on a relatively cheap new cooler, and it helped in the idle temps enormously. Even when turning on the torture test, the temp was rising steadily, but slowly, instead of shooting up instantly like it did with the last cooler.

Unfortunately, on core #7, as the screen froze...
[ERROR: ILLEGAL SUMOUT
Possible hardware failure, consult readme.txt file, restarting test.]
the results.txt has this 22x

I will see if I can turn off the 7th core somehow and run the torture test again to see if that changes things. Will update when I have more info.]


The update:

Well, tried through Prime(using 6 threads, even tried 1), with no luck.
The temperature doesn't seem to be a constant factor after several tests though. Once it froze at 52C after under one minute, and once it sat on 76-77 for about 5 minutes before freezing.
(All done on a clean boot, by the way).

Now I must assume that it is my hardware... is there any way without having a replacement mobo/processor available to test if it is specifically the mobo or the processor?

Regards,
Rebornneo
 
Unfortunately replacement is the only real test you can do with critical components like the board and CPU. The only option (to basically confirm whether they need to be replaced, is to pop the cooler off again, pull the CPU out and double check for bent pins. If there are none, the board probably needs replacement. If there are, try bending them back (VERY CAREFULLY), and then reassemble the system.
 

Rebornneo

Honorable
Nov 17, 2013
5
0
10,510
Thank you for the suggestion, but based on how close after installing the new CPU the problems were caused, I am inclined to blame it. Especially after trying a clean install just to be sure.
As well, the RMA process for the FX shouldn't be long, and the one for my mobo I'm assured is a LOT longer.
If the CPU ends up NOT being the problem, I'll probably get a Surface Pro 2 to pass the time.
Thank you for your suggestions/thinking along.

Unless you don't want me to (because it's not really solved yet, and there's always a tiny chance it's something else), I'll mark your inital response as the solution based on the fact that it lead me to the actual problem.

Thanks so far!
Rebornneo

[Forgot to put this in here: I DID check the pins on the CPU, they were perfect. Still, refer to my thoughts in the first sentence.]
 

Rebornneo

Honorable
Nov 17, 2013
5
0
10,510
Answer found. It was in fact the CPU. A co-worker of mine had another AM3+ cpu in a system he wasn't currently using and borrowed it to me to test this last thing. It's been running all cores(6) for about 20 minutes now at 99%. No freezes/glitches or anything weird. I'm considering this solved.

Thank you The_Prophecy, for the solving suggestion.