2011-02-14, 03:58
Well, I've just built-in my passive Sparkle GT430 in my HTPC for a few days now, and during the testing and observing I did -during the time I have that card now- I noticed some things that I cannot really 'rhyme' with something I would expect.
This is the card btw:
First of all: the card is built-in in an Ahanix MCE301 case, which has zero to none airflow (just one noiseless low rpm 80mm fan). I knew of course where I was getting into with placing such a passive card in that case (so that point can be left out of the equation ).
So what I did expect, was that the temperatures would be a little higher than normally seen on a (passive) GPU. I'm perfectly fine with that. After first testing (menus, playing content etc.) for some short periods of time I never saw temperatures above about 74°C. As expected - almost no airflow, passive card, perfectly fine by me.
But now I have the setup running a little longer there obviously would come a moment that the system would be 'left alone' for a period of time. Running idle so to speak. And indeed - that happened. The HTPC was idling in the main menu for an hour or so before I came back. At that point I decided to check the GPU temperature, hoping to see near refrigerator temperatures, but wow - 84°C ! And that is at normal room temperatures (say about 24°C)! Hmmmm ...!!!
That can't be normal I thought. Displaying some antique OpenGL interface makes the card run hot, but displaying HW accelerated content actually cools it down? So after some more testing here is what I noticed:
System uptime 6.5 hours, so 'stabilized' case environments:
Is *that* normal? I mean: Rendering an OpenGL menu should be peanuts for any card, so why is a GT430 actually running hotter doing that, than when doing its work on HD content?
This is the card btw:
First of all: the card is built-in in an Ahanix MCE301 case, which has zero to none airflow (just one noiseless low rpm 80mm fan). I knew of course where I was getting into with placing such a passive card in that case (so that point can be left out of the equation ).
So what I did expect, was that the temperatures would be a little higher than normally seen on a (passive) GPU. I'm perfectly fine with that. After first testing (menus, playing content etc.) for some short periods of time I never saw temperatures above about 74°C. As expected - almost no airflow, passive card, perfectly fine by me.
But now I have the setup running a little longer there obviously would come a moment that the system would be 'left alone' for a period of time. Running idle so to speak. And indeed - that happened. The HTPC was idling in the main menu for an hour or so before I came back. At that point I decided to check the GPU temperature, hoping to see near refrigerator temperatures, but wow - 84°C ! And that is at normal room temperatures (say about 24°C)! Hmmmm ...!!!
That can't be normal I thought. Displaying some antique OpenGL interface makes the card run hot, but displaying HW accelerated content actually cools it down? So after some more testing here is what I noticed:
System uptime 6.5 hours, so 'stabilized' case environments:
- Displaying content (Blu-ray, mkv whatever): Temp max 71°C
- Switching to menu - idle it for about a short (10 min. maybe more) period of time: 84°C!
- Back to displaying contect, temperatures drop back to about 70°C or even less.
Is *that* normal? I mean: Rendering an OpenGL menu should be peanuts for any card, so why is a GT430 actually running hotter doing that, than when doing its work on HD content?