Setup | Noise | Power consumption | Heat levels

Setting up the PC

Right, at this stage we are going to insert and connect the graphics card into our test setup. Installing the card into your system will be a fairly easy job. Just slide the card into a free PCIe slot, connect the DVI cable to one of the DVI connectors, connect both the 6-pin power connectors to the card.

Especially with a high-end card like this... I do recommend you buy a decent PSU with some reserves, always. The PSU is an extremely important component in your PC. We'll get into that in a minute though.
Once the card is installed we startup Windows. We install our driver, reboot and you should be good to go. The card will work straight out of the box.

The 6-pin power connectors. The tiny connector is the S/PDIF lead.

Power Consumption

On the last page we briefly touched on the topic of power consumption, for a product of this caliber it's really not that bad. In fact on 55nm it even improved a little more.

1400 Million transistors means more power consumption... well you figured that out yourself already I guess. The GTX 200 GPU however is one of the most energy aware products you'll find on the market today. NVIDIA made sure several new power states are active in the GPU. What this means is that depending on the task at hand you assign to the GPU, it'll react to that in terms of power consumption. NVIDIA has integrated 'P-states' and this allows the GPU to alter clocks, voltages and frequencies on the fly. Very interesting is that certain parts of the GPU when not used, can be shut down, conserving energy. In GPU terminology this is called clock-gating circuitry, it effectively "shuts down" blocks of the GPU which are not being used at a particular time, reducing power during periods of non-peak GPU utilization.

I've been trying to figure out a way to actually show these dynamic processes and I can actually show you a little tip of that iceberg.

The GPU for example, the core domain can hop from 300 towards 700 MHz when needed. The shader domain, it's clocked down towards 600 MHz if not used, and can even go as low as 100 MHz if it is not utilized. Once we fire up a 3D application it'll jump up to it's designated clocks, and the same thing happens for the memory.

That's clock gating / P-states. Depending on what level of performance is required by the GPU, it will adjust it's clock frequencies and corresponding internal voltages.

• Idle/2D power mode: approx. 25W
• Blu-ray DVD playback mode: approx. 35W
• Full 3D performance mode: varies - worst case TDP 183W for GTX 285

So though the processor can peak towards 200 Watt with hefty gaming, the reality is that it's just as much as a GeForce 8800 Ultra, yet you have double the performance and the overall picture of this GPU in more common situations seems to be very energy efficient.

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.

Now a couple of things are interesting here. Again look at the graph shown above. You should see the P-states we discussed earlier, normally these will save you energy. In 2D mode the GPU tries to save power by lowering clocks & voltages, therefore running at a much lower core frequency.

At idle you can expect a temperature of 40~45 degrees C. Pretty normal. Yet once we push the GPU to 100%, the temperatures take a pretty hefty toll and settle at 80 to 85 Degrees C. This remains fine within designed operational specifications but it's hot alright.

Though air will be exhausted outside the PC, make sure you PC chassis is well ventilated.

Power requirements

We'll now show you some tests we have done on overall power consumption of the PC.

The methodology is simple: We have a device constantly monitoring the power draw from the PC. We look at the recorded maximum WATT peak; and that's the bulls-eye you need to observe as the power peak is extremely important. Bear in mind that you are not looking at the power consumption of the graphics card, but the consumption of the entire PC. From a performance versus wattage point of view, the power consumption is pretty good with the new 55nm products.

Sidenote: we recently upgraded our test-platform, which by itself utilizes a lot of energy.

It's Core i7 965 / X58 based and overclocked to 3.7 GHz. Next to that we have energy saving functions disabled for this motherboard and processor (to ensure consistent benchmark results).

The ASUS motherboard also allows adding power phases for stability, which we enabled as well. I'd say on average we are using roughly 50 to 100 Watts more than a standard PC due to these settings and added CPU overclock. Keep that in mind. Our normal system power consumption is much higher than your average system.

• System in IDLE = ~200 Watts
• System with GPU in FULL Stress = ~395 Watts

The monitoring device is reporting a maximum system wattage peak at roughly 395 Watts, and for a PC with this high-end card, this is excessive but remains within acceptable levels.

The IDLE Wattage is impressive, though a little higher than the competition. Again, this test PC consumes heaps of power.

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 card 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.

• So in IDLE mode, you will not hear the card. We measure 40-41 dBA coming from the PC.
• Gaming with an average title not stressing the GPU too much we can hear the fan a little, we measure roughly 42 dBA.
• When we loop 3DMark Vantage for a while the GPU really heats up, as a side-effect the fan RPM will go up even higher, the noise-levels are now 43 dBA which definitely can be heard, but it's fine really.

So to sum it up, it's a really quiet graphics card, yet during gaming you will be able to hear it. Nothing intense though.