Testing the beast
A case this accommodating wouldn't be done any justice by using a micro ATX system, so this is the perfect opportunity to introduce our new quad-core test system. We used an AMD 790FX-powered Asus M3A32 MVP Deluxe motherboard paired with a 2.0GHz Phenom X4 9350e. To cool this Socket AM2 processor, we used a popular tower heatsink from Kingwinthe Revolution RVT 9225 HDT with a 92mm fan rated at 28dBa at 2800RPM. An XFX GeForce 8800 GTS 512 filled in on the graphics front, and we threw in threw in two 1GB sticks of Corsair CM2X1024 DDR2 memory for good measure.

In working with the new motherboard's BIOS, I found that the "Q1" smart fan speed control matched up nicely with this heatsink if I used the "Optimal" fan speed setting. With this configuration, the processor cooler generated less noise than any of the case fans at idle. Under load, it spun up just enough to be as loud as the quietest case fan. This time around, we're testing system fans running at a full 12 volts, since the majority of users don't want to have to mess with extra fan speed controllers or voltage hacks. Having fans running at their default speed should make it easier to compare case noise levels, as well.

We chose to leave the power supply area blocked off in all the tests. The bottom of the case has plenty of airflow to cool the power supply, and the heat produced should be exhausted from the case by the PSU's own fan.

In one set of tests, we're leaving the two top divider plates out, so the large side fan will benefit both the CPU and GPU. To really put the Advance Package Spedo through its paces, I also tested the case with all its dividers installed. My expectation was that the CPU might stay cooler with the partitions in, but the graphics card should run hotter since it won't benefit from the generous airflow that should bathe the rest of the system.
The Spedo already has more airflow from the back, front, and top fans than a lot of cases, so I also disconnected the side 230mm fan completely for one test. This should help with noise levels, as I could tell the side fan was producing the bulk of the noise made by the system.
Finally, I've also put the Phenom system into Gigabyte's 3D Mars enclosure to illustrate how the Spedo performs against some competition. We were impressed by the Mars when we reviewed it a few months ago, so it's a good reference point for the Spedo.

At idle and with the Spedo's dividers in place, disconnecting the side fan raises component temperatures slightly. That's not the case with the dividers removed, where we see most component temperatures unaffected by the presence of the side fan. Instead, the temperature of the air around the CPU rises slightly, as hotter air around the graphics card is mixed with the rest of the system.
If you go back and take a look at the 3D Mars' performance compared to a couple of other cases, you can see that it did very well even with its twin rear 120mm fans turned way down to five volts. With those fans now running at full speed, alongside another 120mm fan in the front and two 80mm fans on the side, the Mars still can't keep our test system's CPU as cool as the Spedo. The graphics card, on the other hand, is about the same temperature on both cases.

Nothing really surprising happens when we fire up the rthdribl HDR lighting demo for GPU torture test. The partitions allow the side fan to lower motherboard and GPU core temperatures by a degree, but leaving them out benefits temperatures across the board, especially with the side fan off. The Gigabyte case runs the processor much warmer than the Spedo, but other system temperatures are relatively close, depending on the Spedo's configuration.

After adding a heavy CPU load across all four cores courtesy of Prime95, the value of the Spedo's cooling partitions becomes rather questionable. Turning on the side fan actually increases CPU temperatures without actually making the graphics card run any cooler. In fact, the lowest temperatures come with all the dividers removed and the side fan turned off. Compared to the 3D Mars, this configuration is quite a bit better, especially for the processor.

The quietest way to run the Spedo at stock voltage is to leave the dividers installed and the side fan turned off. Again, though, the sweet spot seems to be leaving the side fan disconnected but taking the top two partitions out, as this is the second quietest configuration and the coolest under load.
The 3D Mars' noise levels put things into perspective. As you can see, the Spedo isn't really a quiet case. That isn't to say it couldn't be made quiet with a few simple modifications. All the fans connect via standard 4-pin Molex plugs, so it wouldn't be hard to feed them 7V or 5V instead of a full 12V. With the sheer number of fans in the Spedo, I'm confident you could achieve any point on the scale from quiet and hot to loud and cold.
| Friday night topic: The trouble with Best Buy | 143 |