Sign in with
Sign up | Sign in
Your question
Closed

Bad luck or something else??

Tags:
  • Nvidia
  • Graphics
Last response: in Graphics & Displays
Share
September 29, 2010 5:42:27 PM

Hello, I have a serious problem. Today my VRAM went bad on my EVGA GTX460. These things happen, but this is the second time this has happened to me with the same model of card.


My question is this. Could something in my system (other that the graphics card) being causing the vram on the graphics card to fail?

I will let you know my system specs and a few other points of interest.

EVGA GTX460 768mb Fermi
Compaq Q2009 Monitor
MSI 870A-G54 Mobo
AMD Athlon X3 440 RANA @ 3.0Ghz
4GB G.Skill DDR3 memory
WD Caviar Black 640GB HDD
Antec ECO 520C - (Antec website says my system requires around 450W minimum)

The 1st time this happened I purchased a XFX HD 5450 so I could use my pc during the lengthy RMA process. This card worked flawlessly for 2 weeks.
I only play World of Warcraft. I leave the pc on at all times and restart once per day.
The monitor has a vga cable so I used the supplied adapter to hook it up via DVI output.
CPU runs at 21c idle and 35-40c while gaming, MOBO stays around 30c at all times, GPU is 29c under idle and 44-50 while gaming.

NOTHING HAS BEEN OVERCLOCKED!

It also appears that this models heat-sink does not contact the vram chips.

Please review these facts. I am going to RMA the card anyways but I do not want this to happen again. Your input is appreciated!

More about : bad luck

Best solution

September 29, 2010 6:01:00 PM

This is usually a sign of power issues. This could be becasue of a bad PSU, power flux in your house line (rolling black outs, power surges, etc.) or an under rated power supply. My suggestion would be to buy a different power supply.

As a side note, why do you restart once per day? Restarting a computer every day just for the sake of restarts is one of the hardest things you can do to a computer. Peak power/power spikes occur when the system first received power. You also have a possibility of loosing data due to incorrect read/write at shutdown. There is a reason that mission critical systems try to remain online at all times and it isn't just for jobs, it because shutdown and startup isn't good for a computer if you don't need to do it.
Share
September 29, 2010 6:25:59 PM

Argh, just checked the surge protector I have and the "building wiring fault" light is on! I have not noticed this because it is behind the desk. Furthermore the surge protector is attached via a 20 foot extension cord. I know the wiring fault issue needs to be fixed but what about the extension cord, are they o.k. to use?

Also it would make sense to me that a power issue exists because the replacement card I used while the 1st 460 was in the RMA process did not have extra power connectors like the 460 does (2 6-pin) and I saw no issues with that one.

I guess I restart once per day out of habit/ignorance, I will not do this anymore.

So I am going to do the following: RMA card, fix wiring fault and buy a new PSU.

Score
0
Related resources
September 29, 2010 6:33:20 PM

supertdubzz said:
I know the wiring fault issue needs to be fixed but what about the extension cord, are they o.k. to use?


No, this is not good practice. So many people use under rated extension cords to run things they shouldn't be. If at all possible remove the extension cord from the mix or ensure that it is a 3 pronged gauge 12 wire extension cord. You could possibly get away with a 3 pronged gauge 14 extension cord but at this point it is better to be safe then sorry. Both the 3/12 and 3/14 are rated at 15A for less then 50' and your house wiring is typically set by breaker or fuse for 15A. There are some cases where you have 20A recepticles but these are far less common and computers are designed to draw less then 15A so as not to trip the breaker or blow fuses.
Score
0
a b Î Nvidia
September 29, 2010 6:38:25 PM

well in the end cooling the vram and the vrm is your task given the choice of brand not that every GTX 460 isn't the same way in one form or the other. I am going to mod mine once I clean it for the first time due to dust build up. I got some vintage low profile vram heatsinks that are like 9 years old. Nice blue heatsinks that cover not one but two vram modules.
Score
0
September 29, 2010 6:55:11 PM

The extension cord is of the 3 prong orange variety and it says 16/3 on it I'm guessing that means 16ga.


I just remembered that I also had a GTS250 on this machine that I RMA'd too because of a fail, but I thought that was related to a different issue.... maybe not.

It has been an eventful 1st build for me that is for sure.

Tx for the help. I will update when I have one!!
Score
0
September 29, 2010 7:00:00 PM

Just to let you know, as wire gauge goes, the bigger the number the smaller the wire and thus the less amps the wire can move. 16/3 is rated at 13 amps for cords up to 50'.

It does sound quite eventful and I hope that your able to fix your problem.
Score
0
September 29, 2010 7:03:53 PM

Best answer selected by supertdubzz.
Score
0
September 29, 2010 10:43:14 PM

So I fixed my wiring fault and the problem is gone, i.e. no artifacts while gaming and EVGA OC Scanner just ran for 10 minutes without finding any artifacts.

Question is, have these issues done permanent damage to the graphics card?

Score
0
a c 173 Î Nvidia
September 30, 2010 2:33:25 AM

This topic has been closed by Mousemonkey
Score
0
!