Is my graphics card not getting enough power?

[J_Adrian]

Hello. I've been experiencing various problems with my PC lately, and I think that this could be related to my video card/PSU, but I'm leaning towards a PSU (hopefully) problem since I changed my generic case (which had 1 80mm fan) to a CM Storm enforcer which has 1 120mm fan, 200mm LED fan, and 2 120mm BitFenix LED fans a week ago.

Here are the specs:
i5-2500
ASUS P8-H61-M LE
Palit GTX 560 OC edition
Seagate Barracuda 7200rpm 1TB HDD
Kingston 4GB RAM
CiVO 1000w PSU (80 Bronze...or so it says)

I can play games for a minute or so before my PC freezes itself. Before updating my drivers, the PC would freeze (even disabling Ctrl Alt Delete), and then I would have to restart it. After installing the latest drivers, the game would freeze for a bit, my monitor would display nothing for a few seconds, the video card fan would slow down (presumably the video card itself restarts, if that's even possible), then I can carry on for another minute or so. Here are some experiences with games/programs I've used:

Before driver update:

DotA 2:
1.) Stutter
2.) Freeze (with sound looping)

HAWX:
1.) Freeze (with sound looping)
2.) Recover
3.) Repeat 1 and 2 a few times
4.) Restart

Crysis 2:
1.) Freeze (with sound looping)
2.) Restart

MSI Kombustor:
1.) Stutter
2.) Freeze (with sound looping)
3.) BSOD (Locale ID: 1033)

After driver update:

DotA 2:
Untested

HAWX:
Untested

Crysis 2:
1.) Freeze (with sound looping)
2.) Recover
3.) Blank screen, Video card fan spin slower
4.) Recover (Video card fan spins on set speed again)
5.) Repeat 1 to 4

MSI Kombustor:
1.) Freeze (with sound looping)
2.) Recover
3.) Stop responding
4.) Display msg (after closing Kombustor): "Display driver has stopped working and has recovered"

 

[RetiredChief]

First PSU is way overkill for a single 560, I'm quessing that your max power consumption is around 350 W. If Power is a problem then PSU falls into the junk catagory.

The first ting I would do is download a program that will display your voltages, primarily the +12 which is the has the highest current drain. Also look at +5 and +3.3. Your Motherboard may have a utility to display these voltages. Alternative programs such as HWMonitor and speedfan, ect.

Monitor the +12 volts and run:
.. Prime95, This will check the CPU and memory for stability.
.. Run Furmark: This will stress the GPU.

.. Both put a High load on the +12 V, Furmark probably the Highest
+12 V should not drop below 11.4 V (I personnaly use 11.6 V as a min)

In Both cases Monitor the Temps!!. Look at the CPU Core temps. With prime95 the CPU will ramp up quickly so monitor CLOSELY for the first 15 Min.

Google a review of your PSU to find out if it falls into the "Junk" catagory and if so - Problem or NOT, REPLACE it.

 

[DelroyMonjo]

Are you using Nvidia drivers from Nvidia.com ?
Lame Q. but I have to ask. You didn't drop a screw while changing cases did you?
No kidding, happened to me many years ago and it killed a stick of RAM and was shorting out my MB to where it wouldn't even get past POST.

 

[Sunius]

Sounds like memory problem. Download memtest and check your RAM!

 

[geekapproved]

What kingston model is your ram, and what settings are you running it at?

 

[sincreator]

Have you tried any different/newer sound drivers for your motherboard? Freezing with sound looping in all the games you play seems like it could be related your sound drivers. http://ca.asus.com/en/Motherboards/Intel_Socket_1155/P8H61M_LECSM/#download

A few months back when I got BF3 I had similar crashes with sound looping. I grabbed the latest sound drivers for my onboard audio and never had that crash again. It was an easy fix. I'm not saying this is your problem for sure, but it's probally a good place to start.

PS: Did you have an ATI/AMD card before the gtx560? If so, did you uninstall the drivers by removing all the drivers and running driver sweeper after?

 

[J_Adrian]

@RetiredChief:

I've tried running prime95 yesterday since I thought it was my CPU, but it's ok. I'll try running both prime95 and Furmark (not at the same time) for 15mins and see what happens. My temps on average are:

Idle (in degrees Celsius):
CPU: 40 ~ 45
GPU: 38 - 40
Hard drive: 35
Motherboard: 35

GPU Load: 75 (I ran until around 83 on Furmark)
CPU Load: 68 (prime95)

Also, I've been trying to figure out why on earth I get weird readings for Cores vs CPU(this is in idle):

I tried stressing the CPU and to my surprise the CPU readings on HWMonitor and Speedfan went down while the Core temps went up. Sensor error/bug maybe?

Also, I've long planned about changing my PSU since all of the people I ask tell me to get a better one. I've been looking at a Corsair GS700.
---
@DelroyMonjo:
I'm using the ones Palit had on their site, and I have no idea if there are screws/etc inside since I was not the one who changed the cases. I can only specify parts while techs do what I tell them (since I have little knowledge of part safety).
---
@Sunius:
I'll check. I haven't thought of RAM since I just got a new one (old one died on me just after 2 months and I'm still on warranty).
---
@geekapproved:

Interesting. I thought RAM was supposed to be on 1333mhz or 1600mhz?
---
@sincreator:
I did have a Radeon HD 5450 2gb, and while I do know that the tech who installed the GTX 560 removed the driver, I don't know if he used driver sweeper as well. The 560 has been in the PC since November last year though.

 

[Sunius]

It's dual channel, means it does it's speed. 667*2 = 1333.

 

[J_Adrian]

Tried updating the sound drivers. Still the same. I also get the "NVIDIA OpenGL driver lost connection with the display driver due to exceeding the Windows Time-Out limit and is unable to continue" error.

Edit 4:28PM GMT+8:
Rolled back video card driver. Same thing.
Removed AMD driver using Driver Sweeper. Same thing.

 

[Sunius]

AMD driver? Why did you have it in the first place?

 

[J_Adrian]

Like I said, I had a HD 5450 2GB before I got the GTX 560. Apparently the tech didn't use Drive Sweeper.
Edit: Ran MemTest and couldn't find any problem there too.
Edit 10:14PM GMT+8: I remember fiddling around with the BIOS settings two days ago since I checked CPU temps, and I tried fiddling with it again today. Turned the settings from High Performance to Auto. Everything seems stable for now and I'm not getting crashes in games, although I still crash on Furmark.

 

[geekapproved]

You have 1333 memory so you should be running the JDEC 4 spec. Are you?

Apparently you don't know, donwload cpu-z, click the memory tab and tell us what it says.

 

[J_Adrian]

You have 1333 memory so you should be running the JDEC 4 spec. Are you?

Apparently you don't know, donwload cpu-z, click the memory tab and tell us what it says.

P.S. I'm terribly new to building PCs (and it shows) so I'm sorry if I can't follow/understand certain things.

 

[Sunius]

hmm memory seems OK. We still didn't rule out the PSU. Do you have a friend that you could borrow a psu from? It might be it after all.

 

[J_Adrian]

hmm memory seems OK. We still didn't rule out the PSU. Do you have a friend that you could borrow a psu from? It might be it after all.

Unfortunately no.

 

[Sunius]

Hmm any chance you could buy a PSU from a shop which lets you return it if you find out it wasn't a power supply at fault?

 

[RetiredChief]

Memory is in Single channel mode (DDR3-1333) - Should add a 2nd 4 gig module, even if you don't need the extra ram. That would inprove memory performance as you would then be running daul channel mode.

On PSU, from what I saw PSU voltages were OK, If that was under Load. However there is a problem with some Low end PSU that even with "good" Voltage levels they can create a problem. This is because the low end PSUs have a higher ripple/noise output and insome cases there are spikes on the rail that are excessive.

On memory test, I find that Prime95 seems to do a better job at than memtest 86 at Identifing memory issue. need to run it for a Min of 4 hours (many leave running longer) - BUT need to monitor temps closely. Do not lete go above 85 C, preferably lower. aproxx 95 C is max for SB CPU.

PS the Stock HSF is at the BOTTOM of the totom Pole interms of performance, Myself I do NOT even take out of the Box. Even a Low cost Hyper 212 is MUCH better at keeping your CPU cool.

 

[J_Adrian]

@Sunius:
I'm not sure, and even if I did, it would take me around 4 to 5 months since I don't get cash allowance on summer. I've thought of replacing the PSU for a long time now though.

@RetiredChief:
I see. I'm going to test using Prime95 again tomorrow.

 

[geekapproved]

Nice catch Chief.

Adrian, you have a dual channel mobo and your only running in single channel mode. In order to get best performance and run dual channel mode, you need to run your memory in pairs.

 

[J_Adrian]

@geekapproved:
So basically my RAM is choking my system since it's only 1 stick running on dual channel mode?
---
Been reading through a thread at the nVidia section and I saw another guy with the same problem as I do, and the guys over there are telling him that the problem is either driver, hard drive, GPU, or PSU. In a few months (or earlier) I'm going to get a Corsair GS700 and probably another 4gb RAM and a new cooler for the CPU/GPU while I'm at it and see what happens.

 

[geekapproved]

Driver? You can rule out driver in 5 minutes, not sure why your still considering that as the problem.

Start with the psu. I've never even heard of Civo and it doesn't even come up on a Google search.

 

[dinkys]

Anly luck? I got Powercolor hd 3870 x2 and have exactly the same problem! In my case i think is the PSU, cause it is from my older Pentium 4 PC and is 405W NeoTech.
I think is the PSU, because the GPU is the only change in my PC for months.