Sign in with
Sign up | Sign in
Your question

CPU, GPU, Memory or PSU?

Tags:
  • Components
Last response: in Components
Share
August 8, 2012 4:55:45 PM

Hello everyone

Just a little detail of my computer spec:

CPU: Intel(R) Core™ i7-3820 Quad-Core 3.60 GHz 10MB Intel Smart Cache LGA2011 (All Venom OC Certified)
FAN: Corsair Hydro Series H60 High Performance Liquid Cooling System 120MM Radiator & Fan (Single Standard 120MM Fan)
HDD: 1TB SATA-III 6.0Gb/s 32MB Cache 7200RPM HDD (Single Drive)
MEMORY: 8GB (2GBx4) DDR3/1600MHz Quad Channel Memory (corsair vengeance)
MOTHERBOARD: (3-Way SLI Support) ASRock X79 Extreme 4 Intel X79 Chipset Quad Channel DDR3 ATX w/ UEFI bios, XFast LAN, Charger, & USB Technology, 7.1 HD Audio, GbLAN, USB3.0, SATA-III RAID, 3 Gen3 PCIe X16, 2 PCIe X1 & 2 PCI
VIDEO: ECS GeForce GTX 550 Ti 1GB 16X PCIe Video Card (2 x GTX 550TI in SLI Mode)
POWERSUPPLY: 800 Watts - Xtreme Gear Standard Power Supply - SLI/CrossFireX Ready

I believe no components are overclocked.

My problem is this... I was playing Diablo 3 then I caught a BSoD. I recently received my computer back from Cyberpower after my other BSoD and it was determined that the harddrive was failing.

I have had the computer back for about a week and starting playing Diablo 3 as near top settings and caught BSoD twice.
I have started the memtest on one of the rams (4gbs) and will be running it for about 12 hrs to make sure

I also started a burn in process for using FurMark for the GPUs and when i stopped it it froze up and BSoD again... temps on both cards when up to 99 degrees! Could this be the issue? How can i test if it is the GPU?

Ive also read that the PSU could be the issue... how do i check this?

or can this be a CPU issue? Please help! I am not that knowledgeable on computer hardware! :( 

More about : cpu gpu memory psu

August 8, 2012 5:12:35 PM

It would help if you could tell us what the error was you were getting with the BSOD. There's an error code associated with every BSOD which helps narrow down what exactly was causing the problem, however getting temps of 99C would be suspect, but then again third party programs can sometimes give suspect sensor readings. Look at speedfan for example. So if you could give some details about the BSOD it'd help.
m
0
l
August 8, 2012 6:34:07 PM

The graphics cards are causing the problem!
That kind of temperatures aren't normal!Are you sure their fans are even working?
m
0
l
Related resources
August 8, 2012 7:09:19 PM

@cl-scott - I didnt get to jot down the error code when i got the BSoD. I will do so if it happens again. Also, I just started the FurMark but didnt do the burn in rather I the idle temps for both GPUs were at the 49 degree to 55 degree celcius mark. When playing Diablo 3 at full settings tempurtures rose to betwen 75 to 85 degrees on both cards. Also ran a real temp at idle the CPU temp is at min range of 30 - 35 and max of 50 to 53

@Kamen_bg i believe temps were rising that much because of the burn in process i started! but i just wanted peoples opinion of what they thought could be the issue.

I will still be doing the memtest later tonight before I go to sleep. but I did however want to get others opinion of the GPU and the power supply i currently have. From what I have read so far the Xtreme Gear 800w isnt that great
m
0
l
August 9, 2012 3:19:25 AM

The card temps are too high... even for Furmark. It's odd both cards went that high. Is the system kept clean of dust and lint? Maybe you need to blow it and the cards out with some compressed air.
m
0
l
August 9, 2012 12:23:36 PM

@clutchc - hm... i just bought the computer so i doubt that there would be any high build up of dust.... also the temps were taken with the side panel of the case open... i thought it would need more ventilation so i purchased 3 additional 120mm fans for my obsidian case... i hope that will cool down things further....

Also, memtest on 1 of the 4gb of ram for 12hrs resulted that there were NO errors! 1 more to go!

Any recommendations of how to cool off the GPU... FurMark scared me so i turned it off! LOL but again it was doing the burn in test or what not.... and you guys are telling that these temps are not normal and is scaring me...
m
0
l
August 9, 2012 3:17:36 PM

Personally, I never run Furmark any longer thanit takes to do the 'benchmark'. If it passes with normal temps, then I don't have any reason to RMA the card. I figure, why shorten the life of the components unnecessarily.

Quote: "Ive also read that the PSU could be the issue... how do i check this? "

I suppose the PSU's Voltage Regulator could be over-volting the cards. The +12V would be the first one I'd check. If you have a VOM meter, test the +12V to GND at the PCIe power connector on the card while the card is being stressed. It shouldn't exceed +12.5V (some even say less). The card also has its own VR, but first see what the PSU is providing.
m
0
l
August 10, 2012 2:46:25 AM

@clutchc - so I performed a bechmark at 1920x1080 resolution for 60 seconds and i got the following stats....

FPS: 15 Max GPU Temp: 94*C THIS BLOWS!!!! Is is this right stats? IS THIS GPU that sucky?!?!

Checked the both 4gb ram sticks and for 10 hrs each and no errors...
I'm thinking its the GPU and or PSU...

Just as an FYI... i have 24 inch dual monitors. Each monitor is hooked up to each of the graphics cards separately... when i play Diablo 3 (as this is currently the only game I have on my PC by GPU temps are at about 85-94 *C!

by the way.. whats GND? and after all the testing is done how can i RMA the card because I dont have original box for it! Cyberpower provided the cards
m
0
l
August 10, 2012 3:08:05 AM

Ah. I didn't know it was a pre-built. No, you'd have to retrun the whole package if you weren't satisfied. Btw, GND is an abreviation for ground*. The PCIe power cable feeding your cards have a colored wire(s), usually yellow and a black wire. The black is (-) or ground and the colored wire(s) is (+) or +12V. Measuring between any of those colored and black wires will give you your actual +12V voltage.

* ground is also referred to as earth in some places.

Yes, don't expect great frame rates when you run Furmark. The FPS will be quite a bit lower than what you will see while gaming. As far as your memtest, it looks like the memory is fine. I wouldn't consider that an issue anymore.

You have your GTX 550Ti's in SLI mode, but each is connected to a separate monitor? When you game, do you use the 2nd monitor as a utility monitor? If so, try this... remove the 2nd monitor from the system. With the SLI bridge in place, just connect the 1st monitor only and temporarily disable dual monitors in Windows. Run the system as if you just had one monitor. Check temps in that configuration. Check back.
m
0
l
August 10, 2012 3:33:21 AM

@clutchc - Yeah.... it was pre-built -=( not too computer saavy here but just know somethings! haha!

As far as frame rates umm... during games I get a bit choppy wish I can improve this somehow

As far as dual monitor... not too sure what you mean by utility... yes it is still plugged into my machine and I am constant monitoring my emails while i game or watch shows etc. etc. when you say disable the 2nd monitor do you mean unplug it or windows settings? If windows, then is there a way I can set it when whenever I play games the 2nd monitor shuts off?
m
0
l
August 10, 2012 3:56:00 AM

By utility I meant just what you described as you use for the 2nd monitor. I wanted to be sure you weren't trying to spread the display across both monitors when gaming.

With 2 GTX 550tis in SLI...
1) You should get better frame rates than 15 in Furmark
2) Low end and mid range cards like yours tend to suffer from a phenomenon known as micro-stutter: http://www.tomshardware.com/reviews/radeon-geforce-stut... It will appear as if the farme rate is lower than it actually is.

If you are trying to use the cards in SLI, but also using one of the cards to feed a 2nd monitor, that could be an issue causing the heat and poor framerate. I would (at least temporarily) not use the 2nd monitor. SLI is suppose to be used to provide a single 'picture' across one or more displays. If you had one monitor or 3 monitors, the display would be spread across the one or all 3 monitors.

Go back to a single monitor setup for now. Disconnect the cable from card to the 2nd monitor and see if things improve. You may have to tell Windows or the Catalyst Control Ceneter to use one display only.
m
0
l
August 10, 2012 4:08:48 AM

i turned off the dual monitor display for a bit and played d3 for about an hour... temps remained at 90*C for 1st card fluctuating to 92 the 2nd monitor fluctuates between 80 to 85. I will try the method and try to use 1 monitor and unplug the other and see if things improve... I will get back to you!
m
0
l
August 10, 2012 4:09:29 AM

Got any pictures of the insides?
m
0
l
August 10, 2012 4:29:44 AM

@amuffin - sure I can provide that but how can i upload a picture here?
@clutchc - been playing D3 the whole time with monitor turned off and unplugged... still having the same temps currently 89C on 1st card and 79 on the 2nd.
m
0
l
August 10, 2012 6:01:27 PM

Is Catalyst Control Center set for a single monitor? Or is it still looking for that 2nd monitor?
m
0
l
August 10, 2012 6:12:08 PM

@clutchc - card is idling at 40*C 2nd card at 54*C... i think its the heat from the 1st card making the 2nd card a little hotter... but it seems that these cards are running fine thus far... i will install my 3 x 120 mm fans when they arrive and the other 2 140mm as well with a fan controller... and see how it goes! I'll update you guys more later!

Also im moved my h60 exhaust to the top middle of my case (obsidian 800d)!

Anyway I can upload pictures to show you my rig?
m
0
l
August 13, 2012 3:13:35 AM

Use a website like imageshack, then copy the link of the photo itself. (under the direct link tab towards the right) and punch in this code:

m
0
l
August 14, 2012 1:02:51 PM

Hey guys just updating on the current status of the computer.. it seems that my computer doesnt get BSoD now but now I get lock ups during game play which makes me hard reset the computer (pressing the power button to restart the computer). Also as I am typing my computer seems it cannot keep up with my keystrokes so I have to type kinda slowly or it'l miss letters.. what gives?Any help would be greatly appreciated.

Also, i started locking up during game play in a mob, GPU Temps were only at 85*C on both...
m
0
l
August 14, 2012 2:48:03 PM

clutchc said:
Is Catalyst Control Center set for a single monitor? Or is it still looking for that 2nd monitor?

Isnt the Catalyst Control Center for AMD cards?
m
0
l
August 14, 2012 6:25:36 PM

nedzad said:
Isnt the Catalyst Control Center for AMD cards?

LOL, you're right. That's what I get when I try to do several threads at once. My bad.
m
0
l
August 15, 2012 12:46:42 AM

Hey guys,

So I tested both my GPu cards one at a time.. and it seems they both run fine. ALONE.

Each card was running at about 76*C to 78*C when playing games

So what I did was I moved one of the GPU down another slot to give them breathing room BUT I cannot connect the SLI bridge...

My question is:

Does it really matter?
What does SLI really provide vs not having SLI bridge?

I'm hoping this will cease the lock ups!

In other news, I downloaded Skyrim and tried playing it and while in the beginning of the game where they are talking I get kicked out to desktop and/or windows restarts... what gives?
m
0
l
!