Freezing Problem on GRID:Autosport with Crimson 17.7.2 Driver

[ao_ika_red]

Hi, everyone.
I have a confusing problem with my not-entirely new PC.
Here's my PC history.
I've built it around a month ago with extremely limited budget and it's my first AMD system ever after years with Intel's (old) C2Q and Nvidia G92. It has 860k Kaveri CPU on Gigabyte A68HM-DS2 combo that I've got just below $80 and on GPU side, I've got Sapphire Nitro R7 360 2GB that also priced about $80. Even though all of them are second hand, but they relatively new with more than 1 yr warranty left on it. Since I've installed OS, drivers, and some productivity software, everything run very nicely. It even manage to crunch 24+hrs straight on seti@home (CPU&GPU @100%).

Okay, now the problem:
I decided to play some games and I chose Grid:Autosport as my first game in this PC. Since I have an old monitor, I decided to use AMD's FRTC locked at 60fps. It run on 1366x768 res with ultra quality. It run very smoothly, but after 15 minutes or so, it freeze. Based on GPU-Z data, my GPU load hovers at 60% but with about 77 C temp. I tried to lower my GPU's peak clock down to 900MHz, but temperature readings almost the same. Is this something to do with the driver or with the game itself?

 

[EXCellR8]

which OS is this?

if only the game is freezing and not the entire PC prompting a hard reboot I would do a clean uninstall of AMD GPU drivers with Display Driver Uninstaller (DDU) and then reinstall the latest WHQL drivers from AMD website. Autosport has pretty low hardware requirements so the game should run without freezing up. i might browse Steam or codemasters forums for reports of similar problems.

 

[ao_ika_red]

I've got the latter problem. The whole system got freeze and the only solution is a hard reboot. I've done clean driver install for several times, but to no avail. I believe there's something wrong with pwm fan controller because my gpu temp will shot up to 70C+ but GPUZ said that the fan speed only about 50%. Unfortunately, I don't have enough money to buy other "modern" games and use it as a comparison. And for now, I'll keep using it as data cruncher (tomorrow is mw@h sprint so there's no room to play on it).

 

[EXCellR8]

In that case, I might test a custom fan profile with Afterburner--which will help keep the temperature of the card down. 70c isn't too much to worry about but by default the speed is controlled by the card's BIOS; I'm guessing that's what GPU-Z reads after you reboot? That would indicate that temps are much higher when the card is working full tilt.

set the fan to 60 or 70 percent and see what that does... if the change allows you to play for a bit longer then we're on to something. if not, and it crashes after just 10 or 15 minutes, we can try something else.

 

[ao_ika_red]

In that case, I might test a custom fan profile with Afterburner--which will help keep the temperature of the card down. 70c isn't too much to worry about but by default the speed is controlled by the card's BIOS; I'm guessing that's what GPU-Z reads after you reboot? That would indicate that temps are much higher when the card is working full tilt.

set the fan to 60 or 70 percent and see what that does... if the change allows you to play for a bit longer then we're on to something. if not, and it crashes after just 10 or 15 minutes, we can try something else.

Thank you for your insight. I've tried several thing this weekend.
First of all, I read my GPU-Z temp while playing, not after reboot. Sometimes I got lucky to play more than 10 min and then check GPU readings.

This weekend, I used my old trusty SpeedFan custom fan setup (I didn't download the afterburner yet to eliminate another potential issue) and set GPU fan speed up to 75%. I play it for a while, and then check the readings (it's 54C + now, so it should be set). But, I noticed that CPU temp was much higher than when I do data crunching (less than 5 C in Overdrive, more than 70 C in SpeedFan). I tried to play in longer time, and then it froze again.

So, I think the real problem maybe the overheating of CPU than the GPU itself. I will replace my TIM and report it later.

 

[EXCellR8]

if you have the space, you could always try a different air cooler... preferably with some copper in place of aluminum. I have a pair of 7870K APU that get pretty hot but space was limited so I used some better cooling and the occasional lock-ups stopped. just another idea...

 

[ao_ika_red]

if you have the space, you could always try a different air cooler... preferably with some copper in place of aluminum. I have a pair of 7870K APU that get pretty hot but space was limited so I used some better cooling and the occasional lock-ups stopped. just another idea...

Yeah, since day one I've got DeepCool Ice Edge Mini FS 2.0 which cost me 10 dollar. Once I've got it run 24 hrs+ doing seti@home w/ sse4.2 inst without any thermal issue (its peak temp only 65 C). But, when I use it for gaming, bad thing happens. My windows error reports told me about display driver reset, Crimson settings reported it as well, hence my first suggestion about driver issue.

I will try reapplying the TIM more carefully, but if there's no difference, I will look for more beefier cooler.

 

[bbhaag]

nvm wrong thread

 

[ao_ika_red]

Update: It's not driver issue nor hardware issue. It's about an antivirus software that somehow flood my ram slowly and cause the system to freeze. I use Avast btw.
Thank you for your advice.