X-bit labs

Workstation Cases Roundup: Best Power Characteristics, Best Cooling, Best Design!

Category: Other

by Yury Vayukin

[ 01/14/2005 | 04:27 PM ]

In this roundup we are going to take a closer look at a few PC case models, which can be used not only as a desktop home system, but also as a good workstation solution. We will introduce to you the products from such well-known case manufacturers as Chieftec, Antec, IN-WIN and HEC Group.

Table of contents:

• Manufacturer: Chieftec Industrial Co. Ltd.
  
• Chieftec CX-01B-SL-B-TC
  
• Chieftec DX-01WD-U
  
• Chieftec BX-02SL-SL-SL
  
• Manufacturer: Antec Inc.
  
• Antec Plus1080AMG
  
• Manufacturer: IN-WIN Development Inc.
  
• In-Win X710
  
• Manufacturer: HEC Group, Compucase Enterprise Company Ltd.
  
• Ascot 6AR
  
• Conclusion
  

This roundup is a detailed study of six system cases that can be used to assemble a workstation rather than a home computer. Unlike a home PC, a workstation must be able to cope with various kinds of tasks, including some narrowly specialized ones, which put one particular subsystem of the computer under stress. This means the internal design of the case, its dimensions and ventilation should be different.

I first give you a description of each case with controls and indicators. After that I stuff the case with the following hardware:

• ASUS P4C800-E rev.2.0 mainboard (BIOS 1016);
• Intel Pentium 4 3.4GHz CPU;
• Intel box cooler;
• 2x256MB Kingston KHX3200/256 DDR SDRAM;
• Sapphire ATI Radeon X800 Pro 256MB;
• Disk subsystem: 1 Maxtor MaXLine Plus II 250GB and 4 Western Digital WD360 36GB;
• 3.5” Teac FD-235HF floppy drive;
• LG GCR-8523B CD-ROM.

I will mention all the positive and negative issues I encountered when assembling the system.

To ensure a stable ambient temperature, I placed the assembled system case into the chamber of a Sanyo MIR-253 incubator that maintained a constant temperature of 25 degrees centigrade inside.

Our tests proper are carried out in several cycles and several steps.

On the first step we measure the temperatures of the key components of the computer when it is idle (the OS is booted up, the power-saving functions are disabled). This step lasts for 40 minutes – this time is enough for the temperatures of the system components to stabilize.

On the second step we are trying to heat up all the system components equally using a prerecorded demo from Far Cry. The duration of this step is 40 minutes, too.

Next, we made two steps instead of one, since our experiments proved it wasn’t appropriate to test the CPU and the hard disk drives under the maximum load independently. So, the S&M utility running at the minimal priority created the load for the CPU, while the HDDs were being heated up by IOMeter (random reading across all four drives). This step lasted for 20 minutes since that was enough for the temperature to stabilize after the previous test step.

I run the first cycle of tests in a system case as it is supplied, i.e. without additional fans. Before running the second cycle I install more fans at the front panel for taking air in. For the third cycle of tests I install exhaust fans at the rear panel of the system case. And during the fourth test step additional fans are installed at both front and rear panels of the case.

I read the temperatures from these sensors:

• Thermal diode integrated onto the mainboard;
• Thermal diode built into the CPU;
• HDD (through S.M.A.R.T.);
• Thermal diode built into the GPU;
• Thermal diode  built into the graphics card.

I used the following software in my tests:

• Microsoft Windows XP Professional with Service Pack 1;
• Crytek Far Cry 1.1;
• S&M 0.3.1a;
• Intel IOMeter 2001.07.19;
• MotherBoard Monitor 5.3.6.0;
• RivaTuner 2.0 RC15.2.

The results are presented as tables or diagrams that show the temperature of the computer’s primary subsystems (CPU, GPU, HDD) in the Idle mode as well as their maximum temperatures with the maximum number of additional fans installed.

As for the maximum acceptable temperature for each particular computer subsystem, it is 68°C with the CPU, according to the manufacturer. The max acceptable temperature of the GPU is not specified, but this core can work at 80°C, according to our information. As for the maximum temperature of the hard disk drives, you can refer to our HDD cooling review to view a graph that reflects the growth of the mean-time-to-failure coefficient depending on the temperature. According to this graph, the possibility of a crash of your HDD doubles at 40°C and quadruples at 55°C! That’s why we think that a temperature of 55°C is not acceptable for a hard disk drive, and the lower the temperature, the better. There’s a slippery moment in our tests and we accept it: we publish the temperature values taken from HDD’s S.M.A.R.T, but the manufacturers each has its own opinion as to where the thermo-sensor should be put. So, each drive reports the current temperature in some unknown spot, and even 70°C may be no cause for panic, for example if the thermal diode is placed on the electronics board and measures the temperature of a chip. A temperature of 70°C is quite acceptable when we’re talking about electronic chips.

In my tests I will compare the temperature of hard disk drives in different system cases, and though the absolute values may be misleading, the comparison of the temperatures will be an indication of how well your HDDs would feel in this or that system case.