Probable Graphic Card Chipset Problem?

TIWH

Honorable
May 29, 2013
102
0
10,690
Hey guys, I just bought this PC and after installing Windows 8 64bit and installing the latest drivers I've been trying to play any game from Assassin's Creed III to Metro: Last Light but after just one minute the system suddenly crashes and the monitor goes NO SIGNAL, I have to mention I've watched around 20 minutes of some of my full HD movies and there wasn't any problem, also the colors and images are fine whatsoever.

Mainboard: Gigabyte GA-970A-UD3
CPU: AMD FX-6100 Black Edition
RAM: Corsair XMS3 4GB DDR3 1600
GPU: Sapphire AMD Radeon HD 7850
PSU: Cooler Master GX550
Monitor: AOC E2243FW2K
 
Solution
Going by the companies track records on quality components I would be inclined to replace the PSU first. Its possible its the GPU but the PSU is higher on my suspect list given the symptoms and whats been done. When you return the PSU, maybe write down or print that list from the eggxpert link and see if they have any 500w or higher that are tier 3 or higher. Also check if they test parts and you could have the GPU tested when you make the trip just incase.

TIWH

Honorable
May 29, 2013
102
0
10,690


A clean one
Alright I'll do that, but don't you think that'd be weird? because I just installed this OS and I haven't even installed half of the programs I use
Yeah all the drivers I got were exactly for W8 64bit
 

TIWH

Honorable
May 29, 2013
102
0
10,690
I just tried running Windows 8's PC Ranking system and the same thing happened, any idea if programs like GPU-Z could make it clear whether this card has a problem or not?
 

bignastyid

Titan
Moderator
Can you try a different power supply the Coolermaster GX(the 450w is the only exception) series are rather poorly built units and have issues with voltage regulation(dirty power) and cannot deliver their rated power (especially in a warm environment). Its very possible the unit is delivering dirty power and/or not enough power and its causing the video card to crash under load(a hd movie isnt going to put alot of load on the video card).
 

clutchc

Titan
Ambassador
All that would tell you is if temp and the core/memory speed was normal. Crashes in gaming are generally (but not always) a driver problem. Are you using the latest AMD driver for the card? Is this a new build? You may want to do a memory check with Memtest to verify you don't have a bad stick of memory.
http://www.memtest.org/
Do each stick individually for at least one full pass.
 

bucknutty

Distinguished
Basics:
Bad or old video driver. Go to AMD.com and get their driver. I never use the driver by the board partner.
Bad ram: run memory check
GPU not seated in the PCI-e slot. Pull it out and reseat it.
GPU or CPU overheating. Use a free program like core temp to check CPU or MSI afterburner to check gpu.
Direct X install is corrupt or damaged. Run checkdisk and check the log to see if there are bad sectors on the HD. Then reinstall DX.

When ever I get a new system or am working a on a system the first thing I do is memory check then check disk. 2 free tools that are built into all windows installs. Start the test and go have a snack. Check disk takes an hour or longer. Memory check only takes like 20 minutes.
 

TIWH

Honorable
May 29, 2013
102
0
10,690


Thanks, I'll check them out, I already got the latest AMD drivers though. BTW any idea about what bignastyid said about the power supply?
 

TIWH

Honorable
May 29, 2013
102
0
10,690
Windows Memory Diagnostics Tool results: no problems were detected
I also run GPU-Z's PCI-Express Render Test and the same thing happened, I managed to observe the GPU temperature at the very last moment before the system crashed and it was around 57C, while the GPU load was %98, so you think that means the graphic card has a problem or should still I go and check out core temp for CPU?
 

bignastyid

Titan
Moderator
Your power supply is crap and could very well be causing your problems since it seems to be happening when the video card is under load and needs more power. Cooler master(also known by some as crapmaster) has a few good units and a few real crappy units, unfortunately you have one of the crappy ones. In this list(http://www.eggxpert.com/forums/thread/323050.aspx) its a tier 4 unit and while there are worse out there I would never suggest getting anything below tier 3. The 650w and 750 w versions of the GX series made the lemon list
http://www.johnnylucky.org/power-supplies/psu-lemon-list.html
 

TIWH

Honorable
May 29, 2013
102
0
10,690


Another moderator told me that about a month ago and I told him this whatever you wanna call it Douchemaster happens to be my best shot in a country where you can't find anything but unknown power supplies and where nobody even knows Thermaltake, Corsair, ... got PSUs, so from what I've told you so far, do you think this current problem is because of the GPU or PSU? if you need more info tell me, I should be taking whatever part that seems to be the problem back to the place I bought it from as soon as possible, since most of the times guarantee means nothing in here
 

bignastyid

Titan
Moderator
Going by the companies track records on quality components I would be inclined to replace the PSU first. Its possible its the GPU but the PSU is higher on my suspect list given the symptoms and whats been done. When you return the PSU, maybe write down or print that list from the eggxpert link and see if they have any 500w or higher that are tier 3 or higher. Also check if they test parts and you could have the GPU tested when you make the trip just incase.
 
Solution

TIWH

Honorable
May 29, 2013
102
0
10,690


Thanks again for the help sir, also the rest of the folks, I'll see what I can do, I think the only other CM PSU they had was the 600w Silent Pro, and it was quite a bit more expensive, now I have no idea if they'd exchange it with that but in case they do so, you think it'd be worth it? (I just noticed it's featured in johnnylucky's evil twin list, the recommended PSUs as well)
 

bignastyid

Titan
Moderator
Thats a much better power supply(tier 3). When you build a computer the power supply is never an area to skimp in. A poor power supply can cause all sorts of issues(bsods, random/unexpected shutdowns, stability issues, and the real bad ones can destroy other components or ven the entire system and usually in a spectacular fashion with sparks, smoke and sometimes fire.
 

TIWH

Honorable
May 29, 2013
102
0
10,690
I just noticed that in Display Settings>Adapter it shows total available memory as 2806MB while my card is the 1GB edition, apparently 1782MB has been shared with the system memory, do you think that's the problem? and is that changeable?
 

bucknutty

Distinguished
Nah that's just the way windows displays the video ram. Something to do with how the driver mirrors the video ram in the system ram so you can alt-tab out of the game or something. On my office computer I have a 8400gs with 256megs of ram. Win 7 says it has 1015 megs of ram. While the NVidia driver says it has 256.
 

bucknutty

Distinguished
I wonder if they have a defective batch. I have the same card in my custom mobile system and it runs great. I go to Lan parties and run it for 5-10 hours of gaming at a time. Never had that issue.

Its in a asrock h77 micro atx board with an i5.
 

TIWH

Honorable
May 29, 2013
102
0
10,690
After checking every single other part (PSU, RAM, HDD) and taking the GPU to the guarantee company I personally checked out every game for more than two hours on one of their systems and it became clear that there wasn't anything wrong with that Sapphire card, the only remaining part was the CPU which the guy told me as I'm using it with AMD's stock fan, that's possibly the problem, today I changed the fan with a better one, the result was that I was able to play Metro: Last Light (1920X1080/4XAA/High settings) with stable FPS (varied around 35-55) for about an hour and a half, but after that it happened again, any ideas?