|
+ Show Spoiler [old post] +Hey TL. I read the Tech Support Guidelines, and proceeded to try the three recommended tests for problems like these: memtest, Prime95, and Furmark.
Specs (all stock):
Windows 7
i5-2500k
GTX 560 Ti
4gb Ram (Corsair brand)
Mobo: gigabyte ga p67a ud3 b3
PSU: 850w
EDIT: 450w <--- thought it was an 850w
All the drivers are updated, including the motherboard BIOS, within the last month. This problem has been occurring for several months.
Now seeing as Prime95 requires 24hours to be effective, I started with Furmark - and it crashed my computer within 5 minutes. No BSOD, just shut down and reboot. My temps were no higher than mid-80's, when Furmark is supposed to run until maximum of 100c, right? What does this mean?
I've been paranoid my motherboard is trash; it had an infinite reboot problem when it came (which we subsequently fixed), but the hanging and freezing rarely if ever results in a BSOD. Unfortunately said motherboard is past its warranty and even got removed from Newegg, so my paranoia isn't quite unfounded.
(and I can't figure out how to get memtest to run, I have no available flash drive and I don't see a handy executable. I've run the built-in windows memory test a few times, but that's always come up clean.)
Such freezing occurs usually in SC2 or Battlefield 3, but I've experienced general system sluggishness and minor hangs in things as little as Google chrome, and it's really bugging me.
Everything's just over a year old. If it's my graphics card, how do I tell? If it's my Motherboard, what are the signs of a dying mobo?
Should I go ahead and try the Prime95 test now? (Though I'm not sure I feel comfortable with it running my CPU at borderline max temperature for 24 hours...)
I think I've included all relevant information; if I missed some, please let me know.
Hey guys, this is an updated version of my first post (above), which no longer accurately reflects the issues at hand.
My system:
OS: Windows 7 64-bit (latest updates)
CPU: i5-2500k
GPU: GTX 560 Ti
PSU: 800w Apower
RAM: 4gb 1600 (running at 1333) DDR3 Corsair brand
Mobo: gigabyte ga p67a ud3 b3
Originally Furmark was crashing, but this was remedied when I realized my PSU was not in fact 850w like I had thought, but actually 450w. I fixed it with my new PSU, but issues are persisting.
The issue currently persisting is this: My computer feels generally slower than it should, with frequent skips and annoying slow downs while doing simple tasks such as browsing and the like. In games such as SC2 or BF3, I will (nearly every game, though noticeably less since putting the new PSU in) experience severe, crippling hangs that (usually) only leave the mouse with responsiveness. These freezes last, in BF3's case, until the game crashes, leaving the system feeling crippled in speed for another minute or two. In SC2's case, I'll be frozen for usually upwards of 2-5 minutes waiting for it to recover. I never crash outright, nor get BSODs; given enough time, the PC just about always recovers from the freezes.
My GPU was open-box, and my motherboard came with infinite-reboot issues (which I later solved). Consequently I'm paranoid of both pieces, but the benchmarks / tests I've run tell me everything is doing just fine. Temperatures were stable and low during all tests, more specifically related in my comments.
Tests / Benchmarks I've Run:
- Furmark: Ran through a half-hour test (as recommended) flawlessly.
- Prime95: Ran through a four-hour test flawlessly.
- Memtest86: Ran through four passes without issue or error.
- CrystalDiskInfo screenshot: + Show Spoiler +![[image loading]](http://i.imgur.com/zyFkB.jpg) (the badass Blastoise is absolutely necessary.)
- HDTune benchmark screenshot: + Show Spoiler +
It's also been noted that TuneUp360 is a "scareware" software with several rogue copies, but the problem has been occurring since before and independent of it, so I'm inclined to think it's not the problem. (However, if you know ways of proving I do indeed have the correct version, I'm all ears. I'd hate to find another problem months from now having to do with this.)
Thank you jaj22 and SkyR for your help up to this point. I hope this updated OP more accurately reflects the information with which we're working with. (Also, apologies for the PSU shenanigans - I had genuinely thought it was an 850w.)
|
Gigabyte offers a three year warranty so unless you bought open box or time machined into the future, the motherboard can't be out of warranty.
The majority of all initial launch boards have been phased out for newer models that are less expensive to manufacture / better.
If your temperatures are borderline max than you mounted your heatsink wrong and you need to remount it. These stress test software puts your components at max load, it does not make it magically go to max temperature unless the heatsink is improperly mounted, poor airflow, etc.
|
On March 31 2012 06:57 skyR wrote:
Gigabyte offers a three year warranty so unless you bought open box or time machined into the future, the motherboard can't be out of warranty.
The majority of all initial launch boards have been phased out for newer models that are less expensive to manufacture / better.
If your temperatures are borderline max than you mounted your heatsink wrong and you need to remount it. These stress test software puts your components at max load, it does not make it magically go to max temperature unless the heatsink is improperly mounted, poor airflow, etc.
Silly me, I made the assumption of a 1 year warranty. Thanks for pointing that out; I think I'll look into getting a new one...
As for the CPU, it idles in the 30's, and when gaming it never goes higher than the 50's, sometimes the 60's. Are you sure it's not Prime's doing? I've got 2 decent fans and the CPU fan is an aftermarket, all in a HAF case and airflow / temperature has never seemed an issue.
Then again I'm not an expert, so please don't take my comment as an insult to your expertise. 
|
Well what are your prime95 load temps? They're probably around 70c which is normal.
|
On March 31 2012 07:15 skyR wrote:
Well what are your prime95 load temps? They're probably around 70c which is normal.
71-73C. I looked up my CPU, and it said the max before "severe degredation" was 72.6C. It makes be a bit nervous is all. (I only ran Prime for about half an hour to check temps, but haven't done the 24 hour cycle.)
|
72.6c is Tcase aka CPU temperature. TjMax is around 100c aka core temperatures.
If you're overclocked, you should reset to factory default to see if it solves your problem.
|
On March 31 2012 07:21 skyR wrote:
72.6c is Tcase aka CPU temperature. TjMax is around 100c aka core temperatures.
If you're overclocked, you should reset to factory default to see if it solves your problem.
I wasn't familiar with the difference. My temp readouts come from Core Temp, so I suppose it's nowhere near my max temp threshold then; good to know!
Everything is stock speed, no overclocking in any areas.
|
Well run prime95 for a few hours and report back with any crashes. You should also get a dvd or usb drive to be able to use memtest.
When you said your computer crashed during Furmark, what does that mean? It shut down, it hanged, it rebooted?
|
On March 31 2012 07:37 skyR wrote:
Well run prime95 for a few hours and report back with any crashes. You should also get a dvd or usb drive to be able to use memtest.
When you said your computer crashed during Furmark, what does that mean? It shut down, it hanged, it rebooted?
It just turned off, then rebooted. When I say it turned off, it immediately shut down, essentially the same way a computer shuts down if you yank the power, the only difference being it started right back up, offered the standard safe mode options, and functioned as normal.
There was no hang prior to shutoff, no BSOD, nothin'.
I'll get on that Prime95 test right now, then.
|
So I've had Prime95 running for 3 hours now, with no crashes or anything of the sort.
|
That sort of Furmark failure is almost always GPU or PSU. Furmark hardly touches anything else.
What exactly are your other symptoms? You mention freezing and hanging but that's ambiguous terminology. Does the mouse pointer freeze? Audio? Do you have to restart the PC to get it working again, or does it unfreeze after a while? Does it ever do this when browsing or just in games?
If you're getting hard freezes and a Furmark reboot I'd be inclined to blame the PSU first. What model is it exactly?
|
On March 31 2012 12:37 jaj22 wrote:
That sort of Furmark failure is almost always GPU or PSU. Furmark hardly touches anything else.
What exactly are your other symptoms? You mention freezing and hanging but that's ambiguous terminology. Does the mouse pointer freeze? Audio? Do you have to restart the PC to get it working again, or does it unfreeze after a while? Does it ever do this when browsing or just in games?
If you're getting hard freezes and a Furmark reboot I'd be inclined to blame the PSU first. What model is it exactly?
PSU: http://www.newegg.com/Product/Product.aspx?Item=N82E16817139009
Specifically it'll hang - freeze and stutter horribly, with something like 0-1 frames/sec, and in games like BF3 or SC2 it'll almost always stay frozen until the game crashes. (SC2 recovers after 2 minutes or so sometimes). The system itself loses all responsiveness, save for the mouse (usually); it'll still retain the ability to move, but alt+tabbing does nothing, and if I can get to another window, nothing there works either. Often causes things like my browser to freeze and crash, skype, other windows I may have open.
Audio continues usually for a short duration following the start of the freeze, but then cuts out. I rarely have to hard-restart the PC to regain functionality; all it needs is a bit of time.
This all goes away once either the game crashes or recovers. Outside of fullscreen games, I'll get random stutters in Google Chrome and sometimes it'll cause crashes, and overall it feels slow in places it shouldn't, like folder browsing or opening relatively small programs (messengers, other stuff). However it doesn't hang with the severity described unless I'm gaming.
I hope that clears up the ambiguity.
I should mention the GPU was an open-box from Frys. You think that's coming back to bite me in the ass?
|
Right, that's definitely not a PSU problem. The symptoms sound most like GPU thermal throttling, which makes sense given the Furmark shutdown. Mid-80s GPU core temp sounds way too low for that though.
|
On April 01 2012 03:20 jaj22 wrote:
Right, that's definitely not a PSU problem. The symptoms sound most like GPU thermal throttling, which makes sense given the Furmark shutdown. Mid-80s GPU core temp sounds way too low for that though.
Damn. So is this enough proof to say that it's definitely my 560 Ti, and that's what needs to be replaced? Could the motherboard cause any of these problems, since everything needs to go through there anyway? (bit of a hardware noob.)
Your description of the problem's possible cause makes sense though. Yesterday I turned the settings on BF3 / SC2 down to the lowest, and I experienced far less hanging or stuttering, with no crashes. If it is indeed thermal throttling, I can see how the high settings on either could potentially get to that point, and why BF3 was crashing every time I got on!
(as a side note, BF3 looks damn beautiful even on lowest settings... lol!)
|
On April 01 2012 05:25 Killcycle wrote:Show nested quote +On April 01 2012 03:20 jaj22 wrote:
Right, that's definitely not a PSU problem. The symptoms sound most like GPU thermal throttling, which makes sense given the Furmark shutdown. Mid-80s GPU core temp sounds way too low for that though.
Damn. So is this enough proof to say that it's definitely my 560 Ti, and that's what needs to be replaced? Could the motherboard cause any of these problems, since everything needs to go through there anyway? (bit of a hardware noob.)
Well, Furmark hardly works the motherboard at all - I've seen it run fine on a machine with known-bad PCI-E. Also running at low settings doesn't tend to reduce PCI-E usage much. It's not an absolute proof, but it's about as good as you get without swapping the graphics card.
Can you get another card to test? SC2 runs OK on medium with some pretty weak/old cards and you're getting very regular problems, so that should be a good test.
|
On April 01 2012 05:50 jaj22 wrote:Show nested quote +On April 01 2012 05:25 Killcycle wrote:On April 01 2012 03:20 jaj22 wrote:
Right, that's definitely not a PSU problem. The symptoms sound most like GPU thermal throttling, which makes sense given the Furmark shutdown. Mid-80s GPU core temp sounds way too low for that though.
Damn. So is this enough proof to say that it's definitely my 560 Ti, and that's what needs to be replaced? Could the motherboard cause any of these problems, since everything needs to go through there anyway? (bit of a hardware noob.) Well, Furmark hardly works the motherboard at all - I've seen it run fine on a machine with known-bad PCI-E. Also running at low settings doesn't tend to reduce PCI-E usage much. It's not an absolute proof, but it's about as good as you get without swapping the graphics card. Can you get another card to test? SC2 runs OK on medium with some pretty weak/old cards and you're getting very regular problems, so that should be a good test.
My roommate mentioned an old (2-3 years) card that I can probably throw in my comp to see how it goes. That and I have the Intel integrated, but I doubt that'll even do SC2 on low lol. So when I switch it out for the other one, if there's no problems in a few hour of gaming, is that pretty much the final proof as to the matter?
Oh and thank you very much for your assistance. You guys also helped me build this rig, so I'm quite indebted
Also, would this explain the slow performance I'm getting while outside of games, i.e browsing and such?
|
Update: It turns out while I was putting my friend's card into my pc to test my system, I found that I don't actually have an 850w PSU. I had thought I did, but it's only 450w. This is most likely the cause of the issue. I picked up an 800w, and as I type, Furmark is running in the background testing, and I'll report back after 30 minutes.
|
Okay, it was indeed the 450w PSU. The problem is no longer occurring - thank you SkyR and jaj22 for your help; despite the fact the problem was not with the GPU, it led me in the right direction and without your knowledge and confidence in the area of issue I'd probably have tried to replace my motherboard.
Apologies for operating on false information.
<3
|
I may have spoken too soon. I seem to be getting the same problems, even though I replaced the 450w PSU with a true 800w PSU. Could the lower voltage have damaged any parts of my PC?
|
Does Furmark still cause restarts?
|
|
|
|
|
|
EPT![[image loading]](http://i.imgur.com/6rRBl.png)
