Our testing methods
Here are the specifications of the Casewarmer inside the 380T:
|Motherboard||MSI A88XI AC|
|Memory||8GB AMD DDR3-1600 (2x 4GB DIMMs)|
|Graphics card||Zotac GeForce GTX 660 Ti AMP! Edition|
|Storage||Kingston HyperX 120GB SSD, Samsung F1 750GB|
|Power supply||Cooler Master V550|
|CPU cooler||Cooler Master Seidon 120V closed-loop liquid cooler|
|OS||Windows 8.1 Pro|
A big thanks to Cooler Master, Gigabyte, MSI, Zotac, AMD, and Kingston for contributing hardware to make this review possible.
I relied on three software tools to test the Graphite Series 380T:
- AIDA64 Engineer for data logging
- Prime95 for CPU torture testing
- Unigine Heaven 4.0 for GPU torture testing
Each test cycle included the following phases:
- 10 minutes of idle time at the Windows 8.1 desktop
- 10 minutes running the Unigine Heaven benchmark
- 10 minutes running both the Unigine Heaven benchmark and the Prime95 CPU torture test
- 10 minutes of idle time at the Windows 8.1 desktop
Since I used the same MSI A88XI AC motherboard as in the Elite 110, I had to go through a similar song-and-dance routine to get a good profile loaded for the CPU fan controller, the details of which you can read more about here. With that accomplished, I proceeded to put the Casewarmer through its paces. The onboard fan controller was tested at its lowest and highest settings, which affected the speeds of the 140-mm intake fan and 120-mm exhaust fan in my test configuration.
The tests and methods we employ are publicly available and reproducible. If you have questions about our methods, join us on our forums to discuss them with us.
Here are the temperatures for the Casewarmer's components, plotted over time:
And here are the minimum and maximum temperatures for each section of the test cycle:
(Note that, due to operator error, I don't have hard drive temperatures for the Elite 110 to compare against the 380T. Sorry about that.)
With the fan controller at its lowest speed, the Graphite Series 380T didn't break a sweat with these components inside. The only slightly worrisome temperature was that of the motherboard, but since this result seems to be following my CPU and mobo combination from case to case, it's probably not fair to blame it on the 380T.
Overall, at the low setting, the 380T's temperatures look pretty close to the Elite 110's—with a couple of exceptions. The 380T's all-mesh side panels appear to be especially good for the GPU. The Casewarmer's GeForce GTX 660 Ti ran 3°C cooler in the Corsair case. The SSD also didn't budge much off its idle temperature, with only a 2°C difference between idle and load. This is excellent performance. I wouldn't hesitate to put even the hottest, most power-hungry components inside the 380T.
I was also curious to see what effect higher fan speeds would have on the 380T's thermal performance. So, I gave the front panel button a couple of clicks to rev things up, and I ran my test cycle a second time. As you can see in the graph above, the motherboard, GPU, SSD, and hard drive all benefited from the added airflow. Their temperatures all dropped a couple of degrees Celsius from my initial test run. I had to double-check my numbers to make sure the flat 31°C temperature of the SSD wasn't a mistake.
While these aren't huge improvements, having the option to increase fan speeds might make the difference between comfortable and worrisome temperatures for more demanding systems. I didn't test the fan controller's middle speed setting, but it's probably fair to say it would perform somewhere between the low and high settings. As we'll soon see, the 380T is well-mannered even with its fans at full blast, but being able to fine-tune the balance between noise and cooling performance is definitely a good feature to have.
While the sonorous roar of an Italian V12 might make for a great soundtrack on the open road, that's not the kind of sound most people want from their PCs. Mercifully, the 380T isn't nearly as loud as its color scheme.
To test noise levels, I relied on an iOS app: dB meter - lux decibel measurement tool. This is just a smartphone app, and you probably shouldn't compare its numbers with the results from a lab-grade decibel meter. However, the app does give us a decent sense of the 380T's relative loudness.
According to the app, the noise floor in my office with no appliances, HVAC equipment, or other computers running is about 30 dBA. To gather my numbers, I took readings from the front, sides, and top of the 380T. I positioned my phone 6" from the case for each reading.
At idle, the 380T runs louder than the Elite 110. Part of the blame probably lies with the all-mesh front and side panels, which don't do much to block noise from the inside of the case. Under load with its fan controller on low, the 380T is about on par with the Elite 110. Kick the fan controller up to high, and the dBA numbers increase accordingly, but not enough to be significantly louder than the Elite 110. The two cases are basically at a dead heat under load either way.
Of course, the character of a sound is just as important as hard numbers. As in the Elite 110, the noisiest components here were the Casewarmer's whiny GPU cooler and the Seidon 120V, whose pump ticked audibly. The stock fans in the 380T were pretty good performers, though. They were nearly indistinguishable from background noise at idle, and even at full speed, they produced something close to an ideal, broad-spectrum "whoosh"—neither unpleasant nor distracting. I'd be fine with having the 380T on my desk full-time, despite its higher sound pressure levels.