My video card keeps overheating and I am NOT overclocking it

[peterwilc]

I recently built a new computer

My videocard is a Sapphire X1900 XT

The rest of my computer, see if this helps.
ASUS A8N-E
AMD Athlon 64 3500+
1 gig DDR RAM
X-Fi sound card
320 SATA 7200.10 Harddrive
480 TRUII power antec
Sviking case -> http://www.tweaknews.net/reviews/sviking/index.php

So yeah... I am thinking it may be my power supply? I know Antec makes quality ones, but maybe 480 isnt enough?

If you think that is fine, then maybe there is a program that will let me set my video card fan speed on full while playing system demanding games?

Basicly, I will be playing Counter-Strike source with the graphics pretty much set on high all around. Get a steady 90+ FPS, then... after about 10 to 20 minutes of playing, this little notice pops up and says "Your Video Card is running at unsafe temperatures, the clock speed has been reduced in order to prevent malfunction"

is 90 degrees C really an unsafe temperature?

I am not overclocking, nor have I ever messed with overclocking.

Any suggestions?

 

[raven_87]

Wierd. Your GPU should automatically (unless overridden) adjust the fan speed to compensate for temps. You might want to try ATI tool and manually set the fan speed higher.

I would like to point out though, your PSU definetly isnt the best choice for powering your system.

 

[peterwilc]

It seems like the GPU fan comes on full power just before it is too late...
I would like to set it to go to full power as soon as I play a game.

And I was thinking about upping to a 550 PSU. How does that sound? Or will I need higher.

 

[DaveUK]

I definitely second the recommendation to maybe try ATI tool to set your fan speed to high on a permanent basis. Maybe try moving the card to a different PSU power connector cable.

But the 480w Antec TruePower is most definitely 110% fine for powering that system, im confused as to why this was even raised as an issue?

ATI recommend a 450w PSU for the X1900XT, and even the lower-end Antec is better than most at that wattage.

I have an Antec NeoHE 550w with a system almost identical to yours (but overclocked CPU and 2GB RAM), upgrade if you want a quieter PSU with headroom for overclocking - but not because you think you need to. The 480w Antec would do you just fine.

 

[peterwilc]

Well, the issue was raised because during demanding games, like Prey, when I would turn on full graphics, after about 10 or 20 minutes, my computer would give these internal "beeps" from within. When ever it beeped, the monitor flashed for a second.

Kind of hard to explain... but it seemed like a warning beep of some kind and I was guessing that something wasn't getting enough juice.

 

[inplainview]

A friend of mine had this problem not to long ago.

Sometimes heatsinks that aren't level and don't make good contact with the chip sneak out.

Either lap (http://www.overclockers.com/tips1015/ the stock heatsink so that it is level and makes good contact or buy an aftermarket cooler for your card.

 

[Granite3]

I have this card in my case, and it gets hot! 80C at load, and after some long sessions at load, would reset back to factory settings, as I used the OCing included with the Catalyst control center.

I cured it with a small oscillating type fan blowing the hot air out of the desk compartment the tower sits in. Has not repeated in the last month or so.

I also replaced the stock cooler with the Arctic Silencer to reduce the noise level, and even with the fan set to 100% I cannot hear it.

 

[slim142]

Open your case, remove the card, make pressure on the heatsink and make sure you put it in the middle of the GPU chip. that way it might get good contact.

If not, then you might have a problem with that fan from the card. Change it before is late.

 

[col-p-todd]

I would say first make sure al the cooling fans in the system are working and if that's done make sure that there is an good airflow in the case. By the looks of it the rear fan is an silent fan which means that that it will generate less air flow but makes it more silent. I would say try replacing it with an Vantec tornado. thy arent quiet but generate a good airflow.

 

[nottheking]

First step: make sure our cooler's working fine. As silly as this may sound, check it for dust, and clean it out if there's any there! For the HSF used for the X1900s, the dust will appear at the deep end of the card, clumped along the copper fins where the air is taken in. then, make sure that the fans are functioning normally; to check, just see if it works well if you use ATiTool to force fan speed to 100% whenever you run a 3D application.

If it seems that the problem is not, indeed, from your video card, it may actually be your power supply; Antec's are rather good, but it may be that somehow, yours was damaged, or for some reason, isn't putting out the power it should. Often, signs of power starvation will appear as over-heating, particularly if, say, the PC re-boots after a while when playing a game.

 

[Slava]

Kind of hard to explain... but it seemed like a warning beep of some kind and I was guessing that something wasn't getting enough juice.

That warning beep is most likely related to your temps. Specifically, your PWM temps (motherboard).

Your video card is overheating because your ambient temperature in the PC case is too high. The card's fan is drawing hot air to cool the GPU. Not good.

1. You need to improve airflow in your case: Use round cables instead of flat ones; see if all your case fans are working right; see if you have enough case fans.

2. Go in the BIOS and check idle PWM temps. If they are higher than 57-59C then under load they can go up to 70C+. This is fine BUT chances are you have some ASUS utility loading at start-up and it beeps as a warning because the standard PWM shutdown temp is 75C.

3. Your motherboard has a PWM heatsink+fan. Check your motherboard manual and see if you can set that fan to always spin at the highest possible RPM (no throttling). This setting may be available in the BIOS. If it is, you should disable throttling for both the CPU and the PWM fans. This will make the fans always spin at the highest possible RPM.

4. It is okay to set PWM warning threshhold to 80C. You can do it in the BIOS.

5. You can disable your ASUS utility which loads at Windows startup.

6. As was suggested above, use ATI tool to set GPU fan to maximum (no throttling depending on temps - cold or hot the fan will spin at its max RPM)

 

[fig7]

I bought a X1900 a few weeks ago. Before I got it I was planning to remove the stock heat sink and fan and water cool the thing. I'd heard that they run very hot so I was worried that my water cooling setup wouldn't be able to cool both the cpu and the X1900 when running at full load.

I was really pleased with the results though. Water cooled the X1900 ran over 20 deg. C cooler than with the original HSF. If was cooler under full load than it was before at idle.

One thing I did notice was that when I removed the heat sink the amount of thermal compund they had used was tiny. It certainly didn't look like enough to transfer all of the heat that the X1900 chip produces.
I reckon that's why it was running so hot before. Once I'd applied some Artic Silver 5 and attached the water block the gpu temps were much more sensible (50 deg. instead of 70+)