Articles: Mobile
 

Bookmark and Share

(20) 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 ]

Temperature and Noise

To transform the barebone system from ASUS into a fully-fledged computer, I added to it a processor, some memory, a graphics card, an optical and hard disk drives. As a result, our system looked as follows:

  • Processor: Intel Pentium 4 3000MHz (Northwood core, 800MHz FSB);
  • RAM: 512MB PC3200 DDR SDRAM by TwinMOS, CL2.5;
  • Graphics card: ASUS V9570/TD (NVIDIA GeForce FX 5700);
  • Hard disk drive: IBM DTLA 305030 30GB;
  • CD-ROM: LITE-ON LTN483-S 48x.

A home computer is likely to meet with two kinds of workload – 3D games and DivX movie playback. So I powered the system up, waited for a while for the temperatures to become stable, then launched the necessary task and wrote down temperature data every five minutes of the next half an hour.

ASUS PC Probe utility showed temperatures of the processor and mainboard, HDTemp examined the hard disk drive and ASUS SmartDoctor helped to monitor the graphics processor and memory on the graphics card. Room temperature was measured with an infrared “gun” thermometer.

First, let’s watch a DivX movie (DivX 5.05 codec with the highest decompression quality).

According to the Task Manager, the CPU workload varied from 20 to 40% during the tests, but its temperature grew by 3°C only. Other system components didn’t heat up at all.

Let’s try something harder. Unreal Tournament 2003 is running in the Team Deathmatch mode 4x4 with seven Masterful bots. Settings: 1024x768 resolution, maximum available graphics quality, full-screen anti-aliasing and anisotropic filtering are disabled in the driver Control Panel:

Unreal Tournament 2003 is a heavy load for both: the CPU and the graphics card – the temperatures of these components grow up. While the processor becomes hot quickly, the graphics card and the mainboard are slower to react.

Again, T2-P shows its advantages: the temperature of the mainboard (i.e. the air temperature inside the system case) only grew by 4-5°C, and the temperature of the HDD remained nearly the same.

Consider the CPU temperature, though. 33°C and 40-42°C in idle and burn modes, respectively, are too low for such a powerful processor, especially with Smart Q-Fan activated and the cooler not working at its full speed. I suspected the utility from ASUS to be wrong and checked the temperature info in the BIOS. The numbers were the same. Then I launched two copies of the BurnP6 utility simultaneously to create an extreme workload onto the processor, but the maximum temperature remained the same 40-42°C.

I have only one explanation of such strange results: the hardware monitoring circuitry and the BIOS of the P4P8T mainboard use a diminishing coefficient when showing the temperature data from the thermal diode. I don’t know if it is a flaw or a deliberate trick.

Anyway, the rest of the thermal diodes agree that Terminator 3 provides a good thermal environment for the system components.

Terminator 2 from ASUS is less noisy than many other SFF PCs, as it is free from the main noise-maker – the loud fan in the PSU. The big fan in the PSU of Terminator 2 works much quieter than small, but high-speed fans in PSUs of other systems.

Another advantage of Terminator 2 is its advanced control over the rotation speeds of the cooler and the exhaust fan.

The only drawback of Smart Q-FAN technology comes from its key principle: the CPU temperature never remains absolutely constant, so Smart Q-FAN is always adjusting fan speeds a little: it’s harder to get used to noise that’s changing its tune all the time.

The main source of this noise is the processor cooler from Thermaltake, but you can replace it with a less noisy model.

 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 ]

Discussion

Comments currently: 20
Discussion started: 03/03/04 11:50:21 AM
Latest comment: 02/14/08 05:29:18 AM

View comments

Add your Comment