GPU temperatures | Noise levels

GPU temperatures

So pretty much once we fire off a hefty shader application at the GPU and start monitoring temperature behavior as it would be mid-gaming, we literally stress the GPU 100% with our test. We measured at a room temperature of 21 degrees Celsius.

On average at idle you can expect a temperature of 40~45 degrees C / 111 F. Pretty normal. Once we push the GPU to 100%, the temperatures take a pretty hefty toll and jump over 80+ Degrees C ~ 179 F. This remains fine within designed operational specifications, but it's quite hot alright.

For cards with reference coolers, air will be exhausted outside the PC, but make sure your PC chassis is well ventilated.

What surprised us was the Palit card. It comes with custom cooler, two fans and a heatpipe based design. Despite the fact it's clocked at reference speeds them temps were among the highest measured.

Here you can see how we measure -- this is the reference GTX 275 card by the way.

Noise Levels coming from the graphics card

When graphics cards produce a lot of heat, usually that heat needs to be transported away from the hot core as fast as possible. Often you'll see massive active fan solutions that can indeed get rid of the heat, yet all the fans these days make the PC a noisy son of a gun. I'm doing a little try out today with noise monitoring, so basically the test we do is extremely subjective. We bought a certified dBA meter and will start measuring how many dBA originate from the PC. Why is this subjective you ask? Well, there is always noise in the background, from the streets, from the HD, PSU fan etc etc, so this is by a mile or two not a precise measurement. You could only achieve objective measurement in a sound test chamber.

The human hearing system has different sensitivities at different frequencies. This means that the perception of noise is not at all equal at every frequency. Noise with significant measured levels (in dB) at high or low frequencies will not be as annoying as it would be when its energy is concentrated in the middle frequencies. In other words, the measured noise levels in dB will not reflect the actual human perception of the loudness of the noise. That's why we measure the dBA level. A specific circuit is added to the sound level meter to correct its reading in regard to this concept. This reading is the noise level in dBA. The letter A is added to indicate the correction that was made in the measurement. Frequencies below 1kHz and above 6kHz are attenuated, where as frequencies between 1kHz and 6kHz are amplified by the A weighting.

We test the cards on dBA levels. Obviously the reference coolers all perform roughly the same. The customized coolers are either louder or softer depending on what they try to achieve.

The noise levels are based in a stressed environment, meaning the GPU was (overloaded) with a shader application.

So the Palit card is the only one with a custom cooler. All others use the reference cooler, and though you can hear them when the GPU gets really hot, they are fine when it comes to noise levels. The Heatpipe based Palit GTX 275 however performed the best here and was 2 DBa more silent. The 40 DBa noise level is a threshold of where you actually start to hear noise or not, at least for me.

So while all cards are good and fine, the Palit did shine here, yet unfortunately also had the highest temperatures.

So to sum it up, the reference cooled models are relatively quiet graphics cards, we like that very much.