Khali wrote:I have few questions that might help the peeps around here help you out a bit faster.
What version of windows are you running? What temps are your CPU and GPU running? Overclocked CPU and/or GPU? Have you updated the bios on the motherboard?
I feel that i experience more blue screens and application crashes than I should.
cheesyking wrote:I feel that i experience more blue screens and application crashes than I should.
You should never get a BSOD so the fact that you're getting any suggests something bad is going on. Perhaps if you could post some of the error codes you're getting someone may be able to help you. Have a look through your system event logs for errors and let us know what you can see.
This is purely a guess but I'd suspect either a dodgy driver or something hard drive related (since memtest didn't find anything wrong with the ram) but this is purely a guess.
EDIT: don't bother posting masses of stuff just yet, the basic descriptions of the errors (if there are any) should do for now.
Chrispy_ wrote:Type eventvwr into the start menu for the eventlog; You want Windows Logs >> System
Sort by either date/time and look for the period when it lags out, or sort by type and check for any "error" or "critical" events during periods you remember being bad.
Can you also run Furmark for half an hour (or until your temperature graph levels out); 92 degrees C is very hot indeed. Most graphics cards have fan controllers that aim for about 80 degrees at most. Usually they're tuned to never hit 90 degrees as prolonged running at those temperatures reduces the lifespan of the GPU.
If it's not sounding like a hurricane in a box at 92 degrees then there's either a problem with the temperature sensors, or a problem with your graphics card's cooling. This review of your card seems to indicate that it runs about 65 degrees under load, so at 92 degrees I'd expect the fans to be screaming at 100%
Chrispy_ wrote:The general sluggishness and BSODs point to a problem with your SSD or hard drives. Could be the drive itself is failing, or the cable is bad, or the controller on the motherboard is bad. As Chrispy_ noted, checking the system event logs for anything weird is a good idea.
If those GPU temperatures are accurate, you also have a GPU cooling problem. So either your GPU temp sensors are messed up, your case ventilation sucks, or the GPU cooler is bad.
The CMOS reset thing doesn't make sense though. If a CMOS reset temporarily cures the problem, that points to some sort of motherboard issue.
You've got a lot of different symptoms that are pointing to different things here; this could be a bitch to troubleshoot.
ryko wrote:gpu temps are a good place to start. try running with your side panel off to see if it makes a difference. vacuum out your case and clean your videocard heatsink, maybe even re-seat the heatsink and re-apply thermal paste. and have you tried replacing the CMOS battery with a fresh one yet?
i would also check your ssd life and endurance rating, etc...with the manufacturer's supplied toolkit first then other utilities like ssdtweaker. in your case, your samsung 830 needs the samsung magician software which can be found on this page. http://www.samsung.com/global/business/ ... loads.html
also you may want to do a firmware update if you're not running the latest version. info available on that download page.
last, you could re-install your OS to one of your wd black hdds instead of the ssd. I have seen ssd failures/deaths have strange effects like the stuttering or lagging you describe.
ryko wrote:well first get a new CMOS battery...they are cheap. 1 for $5 locally or 10 for $5 on amazon. or at least use a digital multimeter to get an accurate reading of the current battery's voltage to see if you need a new one. the clear CMOS temporarily making things better could point to a bad battery.
very strange that temps don't change at all with or without the side panel. what is your ambient room temp? try cpuid hwmonitor and maybe try putting a box or room fan next to it while stress-testing the gpu.
but my gut tells me it is your ssd. if you don't want to install your OS on one of your wd blacks then find/get another hdd/ssd to test as a boot drive. ssds don't always fail gracefully and with predictable results. was the drive hammered with writes during it's lifetime? was it trimmed regularly? i know windows 8 uses some other euphemism for trim... see here: http://www.thessdreview.com/daily-news/ ... h-comfort/
and more info: http://www.lancelhoff.com/how-to-check- ... s-enabled/
Chrispy_ wrote:Something is seriously wrong with your temperature results. You're getting 70 degrees over ambient with 100% fans and a honeywell pointed at the graphics card. Anything north of 60 degrees in that situation would be a cause for concern, especially with the oversized cooling on your GTX680.
I'd be inclined to say that it's not really 91 degrees, but that you have a software/hardware temperature sensor fault; This could cause all manner of unnecessary throttling.
I've never come across a graphics card that doesn't get obviously noisier when Furmark is run.
I'd remove and reinstall the Nvidia drivers, and also reinstall motherboard drivers - chipset drivers in particular
Remove anything to do with RoG or Asus Probe - Asus' monitoring software may be doing something freaky, and you can always reinstall it later if the issue turns out to be something else.
grapenuts wrote:The motherboard drivers on asus's website are listed as beta for windows 8 which is why I didn't bother installing them and left the default ones that installed with windows 8 (nothing is missing from device manager.
Chrispy_ wrote:grapenuts wrote:The motherboard drivers on asus's website are listed as beta for windows 8 which is why I didn't bother installing them and left the default ones that installed with windows 8 (nothing is missing from device manager.
I'm not sure how important they are but at a bare minimum you want the intel chipset drivers (you can get those from Intel directly) and potentially stuff like the Intel Management Engine drivers. I'd start by installing the beta ones from the Asus site, but I wouldn't trust Microsoft's generic W8 driver to get everything right.
There's a reasonable chance you have faulty hardware though. The fact you can run your graphics card at full load without the fans even spinning up is NOT right.
Phaleron wrote:Can you install nirsoft bluescreen viewer and copy and paste the results? edit: it should point us at the the driver/piece of hardware that is at fault (hopefully I'll have internet to check back quickly)
edit: for clarification
Arclight wrote:Try to do the following, but be carefull:
- download and install MSI afterburner, get into the options and tick the options to show you the GPU temps, GPU load and fan speed while ingame;
- open the side panel of the case and ground yourself;
- start a GPU intensive game and as the temps reach a stable high value gently press the GPU cooler up against the card around the area where the GPU is located, usually in the middle of the card and see if the temps get slightly lower.
If the temperature gets lower then you have a problem with the heatsink and since it's under warranty it would be advised to contact the company that sold it to you and ask for a replacement.
The system becoming unresponsive for a minute or so could be caused by the video card drivers. It's possible you could be getting one of the infamous Timeout Detection Recovery errors without getting the pop-up warning (happened to me with one of the recent official drivers but i find out from the Event viewer error).
Next time your system lags look at the time, wait for it to recover, go to the Control Panel, Administrative tools, Event Viewer and click on Administrative events on the left side of the panel. You will have to wait for a bit for the errors to be displayed, when they do look for one that happened at the time the system started to behave strangely. Click on the error and google the event number and what other information is displayed.
Khali wrote:If it was me I wouldn't be satisfied with anything over 70c. Your card is going wear all that much faster due to all that heat. Sounds to me like its time to try removing the heat sync and cleaning the old thermal compound off the processor and applying a new layer. You might consider going out and getting some small washers to put on the heat sync screws so you can tighten them down just a little more. Just don't over do it and strip the threads.
If the card is still under warranty I would seriously consider trying to get a replacement. This one has been abused due all that heat and is likely to fail a lot sooner than it normally would. This is the reason I pay a little more and get EVGA cards, they never quibble over sending out replacements.