AOpen XC Cube EZ65: A Quiet, Powerful SFF

Complete|Mobile Systems
Viewing page 4 of 6 pages. Previous 1 2 3 4 5 6 Next


The following components were tried in the AOpen XC Cube:

Two different 7200 RPM hard drives were tried.

  • Hitachi / IBM Deskstar 7K250 250G Serial-ATA hard drive (HDS722525VLSA80). This is Hitachi's highest performance non-SCSI model, a 8 MB cache, 3-platter design;
  • Samsung SP0802N, very similar to the previously reviewed SP1604N in performance and noise (as per the link), a modern, higher performance successor to the legendary Seagate Barracuda IV, whose level of low noise is no longer matched by current Seagate drives. This drive was used to establish the lowest noise level the AOpen XC Cube is likely to provide for most users.

Microsoft Windows XP Pro SP1 was installed, along with the multi-megs of updates ad nauseum. A drive image was used to have exactly the same setup on the two drives.

The assembly went very smoothly, with the logical open layout, and the multiple photos and step-by-step instructions provided in the manual. The basic approach is:

1) Remove the drive cage. Install the drives in it. The external drives require depth adjustment to line up just so in the case, so leave them unbolted for now.

2) Install the memory sticks and any necessary data cables for drives, all very easy with the drive cage removed.

3) Remove the CPU cooler, install the CPU. There is a waxy layer of thermal interface material on the base of the CPU cooler, similar to what Intel uses on their stock heatsinks. I chose to remove this in favour of Arctic Silver 3, applied as per AS directions. Then the heatsink was installed - a very simple task.

4) Slide the populated drive cage in place, adjust the depth of the optical drive and secure down with side screws, then insert the two top screws to lock the drive cage down.

5) Insert cables for power and data as necessary. This was made super simple because of the excellent cable management -- everything was already right there in there right place. The only cable I had to rout was the SATA drive cable, which was looped around the drive to the front. A little shaping of the somewhat stiff cable friction-fit it nicely into place, then all was set to go.

As you can imagine from the above description, it was a quick and painless procedure. AOpen has to be commended for their excellent layout, cable management and friendly manual. Almost anyone who can handle a screwdriver can probably handle this assembly.

BIOS Flexibility

The XC Cube EZ65 continues a general AOpen appoach to their motherboards: Very broad user options in the the BIOS.

  • Vcore: 1.1~1.85V
  • FSB: 100-400 MHz, in 1MHz increments
  • AGP clock adjustments in 1MHz increments
  • Wide range of AGP and RAM voltage and clock options
  • Flexible memory clock options

This is a full-blown enthusiast style BIOS with a huge range of options for all the parameters you could wish for. The dual channel SATA drive support and RAID option is the icing on the cake for performance enthusiasts.

One Oddity

When the cooler was removed to install the CPU, I discovered an oddity: A film temperature sensor in the CPU socket. Such devices were used in AMD motherboards when the Athlon first appeared. They did not have an embedded thermal diode in the die, so motherboards were equipped with these external thermal sensors. (An unsatisfactory problem because the external casing temperature is not the same as the core temperature, and the relationship between them is not linear.) In any case, a temp-diode was integrated into the AMD XP chips, and current generation AMD-socket motherboards do not feature the external socket temp sensor.

The Intel P4 was equipped from day one with an internal temp diode, so why this AOpen board features an external sensor in the CPU socket is a mystery. I could be mistaken about what the thing is... AOpen did not get back in time with an explanation before this article was posted.


Two configurations were tried:

1) Value/Low Noise: P4-2.53 + Samsung SP0802N. Approx market price: US$710

2) High Performance: P4-2.8C + Hitachi 7K250 HDD + nVidia Ti4800 VGA card. Approx market price: US$1070

Much of the price difference is in the hard drive and AGP VGA card: The Samsung can be had around for $70, while the Hitachi is typically around $270. The Ti4800SE card can be had for $140. There is only a $25 difference between the CPUs, if you can find a P4-2.53. This model was one of those discontinued by Intel in their latest price announcement. (Shocking as it may seem for those who remember Intel's struggle just a couple of years ago to break past 1GHz with their P3s, the 2.8 is now the the slowest P4 they produce.)

The reasons for trying these configurations were simple:

  • SFF packages are perfect high value systems, with their built-in everything and low start up cost. It probably the way a large percentage of these systems are used. It is also the easiest way to set up these systems, definitely a consideration for the less technical market these packages appear to be focused on.
  • On the other hand, the AOpen XC Cube EZ65 appears ideal for a portable LAN-gaming type machine with the Intel 865 chipset and the impressive range of performance features, so this seemed worth exploring. It was also interesting to see how much of a hit noise performance would take in such a configuration.

Undoubtedly, some will complain that the Ti4800 is not tops for performance. No it is not, but the reality is that you would not want to put a current top performance VGA card in most SFF. The space restrictions are such that cooling will be an issue with VGA cards pumping 60W of heat and more. The Ti4800 also runs rings around ANY video card integrated into a motherboard; it serves well enough for our purposes.


These comments refer to the following configuration:

  • Intel P4-2.53 CPU (533 MHz FSB)
  • Samsung SP0802N HDD
  • Integrated VGA on XC Cube
  • OCZ PC-3200 256MB EL DDR Platinum Dual Ch. SDRAM Memory x 2

As with the Shuttle Zen XPC, the hard drive added a significant humming noise to the system. A simple suspension with elastic cord was tried, with results similar to that achieved in the Shuttle: Virtual elimination of the humming noise.

A. Power, Temperature and Fan Speed

The AC power draw was measured using the Kill-a-Watt AC power meter. The other data in this table was obtained with the AOpen SilenTek thermal fan control utility. Motherboard Monitor could not be configured to work properly with this system. The ambient temperature was 20~21°C during testing. SilenTek was engaged in manual fan speed mode with the CPU fan set to 40% speed.

AC Power
CPU Temp
Sys Temp
HDD Temp

All temperatures in °C

The accuracy of the temperatures reported are in question. The CPU and System temperatures seem too close, and it's odd that SYS temp is sometimes higher. One wonder whether System temp is really the CPU internal diode temp... and the CPU temp is the readout from the in-socket temp sensor? Hopefully, AOpen will tell us.

In any case, none of the numbers are cause for alarm, the power draw is very modest, as are the reported temperatures. The HDD temps are surprisingly low, considering the tight space. Heat conduction to the aluminum chassis must be a factor. The fan speed ramp up is audible, but not much. More on noise later.

B. Performance

Benchmark tests were performed using PCMark 2002, 3DMark2001 and Sisoft SANDRA -- as done by Ralf Hutter in the Shuttle Zen XPC review. The only difference here is that a newer version of Sisoft SANDRA was used. The raw numbers from those benchmarks are thus not directly comparable, but the same comparative references are provided in each test, which should help keep things in perspective.

The CPU score here is directly related to clock speed. The HDD score is substantially higher than in the Shuttle Zen XPC review; this is a direct reflection of the performance difference between the Seagate Barracuda IV and the Samsung SP0802N.

You knew this one would be bad. It's half the score achieved by the Shuttle Zen with its ATI 9100 chipset, thanks to Intel's "Extreme" Graphics. As one reviewer once wrote, I probably misquote, the only thing extreme about it is how extremely slow it is.

No surprises.

It's dual channel memory as with the Shuttle, but the system is at 533MHz FSB instead of 800MHz, hence the lower score.

This config performs very similarly to the Shuttle in the multimedia benchmarks.

C. Noise

The AOpen XC Cube at idle in this configuration is as quiet as claimed. Its level of acoustics in normal usage is close enough to the Shuttle Zen that switching back and forth between them, it's difficult to make any clear choice about which one is quieter or noisier. There may be no real difference. I'd call it a tie. Both are very quiet, not quite at the same level as many silent enthusiasts' home-brew machines, but certainly better than other SFF PCs and most commercial production PCs.

With a simple elastic cord suspension of the hard drive, the overall noise comes down to excellent low levels. The improvement from the HDD suspension is subjectively much more significant than seen in the SPL measurements below, because the more irritating hummmm noise is pretty much gone, and most of the remaining noise is the more benign whoosh of wind turbulence.

Mic 1m from
Standard HDD Mount
Suspended HDD
HDD Peaks
HDD Peaks

* All measurements are in dBA @ 1 meter
* Max is measured after >20 minutes of CPUBurn stress testing, which heats the CPU more than any other program or utility we know of.
* HDD peaks is the peak noise during hard drive defragmentation.

Where the AOpen gets a bit louder than the Shuttle is during prolonged CPU stress testing. It does not appear to be the CPU fan that's responsible for the bulk of the noise increase, but mostly the fan in the power supply, obviously speed-controlled by temperature. The increase in noise is 2~3 dBA on average, noticeable, and with the HDD mounted normally, brings the overall noise to the upper borderline of quiet. With the drive suspended, however, the acoustics stay pretty low.

In contrast, the Shuttle does not have this problem, not going up in noise level even under prolong stress. (Note that the Shuttle review does not feature a Max column for SPL because its CPU fan speed did not change under load.) This is because the Shuttle's PSU is external and fanless, and its heatpipe cooling system may be a bit more efficient at removing the CPU heat while staying at the default fan speed.

Previous 1 2 3 4 5 6 Next

Complete|Mobile Systems - Article Index
Help support this site, buy from one of our affiliate retailers!