Quiet Mini-ITX Gaming Build Guide #3: BitFenix Prodigy Edition

Do-It-Yourself Systems | Silent PC Build Guides
Viewing page 6 of 7 pages. Previous 1 2 3 4 5 6 7 Next

BASELINE NOISE

Before any stress testing let's take a quick look at the noise produced by each part. We did this by leaving the system on but idle, and stopping/unplugging all the fans (except for the power supply) to isolate individual components.

GTX 980 + BitFenix Prodigy System
Component Noise Levels (idle, power supply on)
Noise-producing Components
Fan Speed
SPL @1m
CPU fan
500 RPM
13~14 dBA
650 RPM
15~16 dBA
800 RPM
19~20 dBA
1000 RPM
26 dBA
Case fans
500 RPM
13~14 dBA
650 RPM
14 dBA
800 RPM
16~17 dBA
1080 (max)
23 dBA
GPU fans
730 RPM (min)
16~17 dBA
1000 RPM
21 dBA
1200 RPM
25 dBA
Measuring mic positioned 1m at diagonal angle left/front of case.
Baseline noise level (PSU on only): 13~14 dBA@1m Ambient noise level: 10~11 dBA@1m.

Our ultra-quiet hemi-anechoic chamber has a noise floor of 10~11 dBA@1m, and the baseline of this system, with just the PSU on, measures 13~14 dBA@1m. The power supply plus either the CPU fan or the two system fans running at 500 RPM produces the same result, so both are apparently quieter than the PSU itself. However, at higher speeds, the CPU fan starts to generate more noise than the two case fans combined. The GPU fans actually don't spin up at low temperatures, so the card is completely silent much of the time. In conjunction with the PSU fan, they emit 16~17 dBA@1m when running at 730 RPM, the starting speed. Once a heavy graphical load is applied, the video card quickly becomes the loudest component.

Of the three sets of fans we're focusing on, the GPU fans are easily worst sounding. They produce multiple tonal peaks in the 200~900 Hz range, giving off a somewhat harsh, grinding type sound, so the video card fans should be kept at as low a speed as possible. The case fans are a bit of a mixed bag, with the sample at the front sounding mostly benign while the sample at the rear seems to be faulty as it generates a bit of a rattle at higher speeds. If the exhaust fan needs to be sped up during the course of testing, depending on the overall sound, it may need to be replaced to maintain acoustic integrity. The CPU fan on the other hand, has a relatively smooth acoustic profile and any negative qualities its saddled with aren't audible from outside the case.

STRESS TESTING: CPU-centric

Testing begins with CPU-centric applications to see how it performs with non-gaming tasks. CPU and system fans are set to a mere 500 RPM, a speed which barely has an effect on the overall noise level, and the GPU fans are turned off as the automatic control dictates. The machine measures just 14 dBA@1m which is just barely audible in our anechoic chamber.

System Measurements
System State
Idle
x264 Playback
Video Encoding
Prime95x4
CPU Temp
30°C
30°C
39°C
44°C
MB Temp
31°C
31°C
36°C
39°C
GPU Temp
37°C
38°C
39°C
40°C
GPU VRM
39°C
44°C
41°C
42°C
System Power (AC)
36W
41W
84W
102W
CPU and system fans at 500 RPM. GPU fans off.
Ambient temperature: 22°C.

The system runs quite cool as one would expect, with CPU and GPU temperatures staying below the 40°C mark on lighter workloads. Video encoding and Prime95 creates much greater demand on the processor but only produces a CPU temperature rise of 9°C and 14°C respectively for the two tests. The board temperature experience moderate increases as well but the GPU for the most part is unaffected.

STRESS TESTING: GPU-centric

For our GPU-intensive states, we use the Resident Evil 6 Benchmark Tool which benchmarks for a couple of minutes, displays the result, pauses for a short interval, and repeats. We tried a few other games but this one seemed to put the most stress on the GPU. The other test is a more demanding combination of Prime95 and FurMark, an incredibly stressful utility that pushes the GPU to its limit. For this state, we run Prime95 with only two threads instead of the maximum four, as most games run with less than 50% CPU utilization. The two combined still produces much more heat than any PC game title. The ASUS STRIX GTX 980 has a fan override that comes into effect at a GPU temperature of about 90°C so we're shooting for 85°C or lower as our target.

System Measurements
System State
Prime95x4
Resident Evil 6 Benchmark
Prime95x2 + FurMark
GPU Fan
off
730 RPM
880 RPM
CPU
44°C
52°C
55°C
MB
39°C
48°C
50°C
GPU
40°C
84°C
85°C
GPU VRM
42°C
106°C
114°C
SPL@1m
14 dBA
19 dBA
20~21 dBA
Power (AC)
102W
260W
278W
CPU and system fans at 500 RPM.
Ambient temperature: 22°C.

Despite more than doubling the system power draw compared to a full four thread blast of Prime95, the Resident Evil 6 test doesn't prove to be much of a challenge. The modest 500 RPM CPU/system fan speeds are still adequate and the minimum GPU fan speed is enough to keep the graphics card under the target temperature. The overall SPL does spike by 5 dB, with the GPU fans contributing most of the extra noise. The Prime95 + FurMark test is much more stressful, requiring an GPU fan speed increase of about 150 RPM, just enough to exceed the 20 dBA@1m level. Even though neither of these two states fully load the CPU, the processor was noticeably warmer than the Prime95-only test, a result of all the heat coming off the graphics card.

At this point we tried to tweak the fan speeds to produce better overall performance but modest increases resulted only in slightly better temperatures and ultimately we weren't able to drop the SPL while maintaining the same GPU temperature standard. Popping off the dust filter off the top of the case had a similar effect, giving the heat a less restrictive exhaust point but its presence helps mask some of the noise being produced. The measured noise level does change depending on where the mic is positioned in relation to the system though, as the main noise producer, the GPU fans, are located on the left side. Angling the mic more toward the front of the case can get it down to 20 dBA@1m, while aiming it straight-on to the side panel bumps it up to 22 dBA@1m.

We mentioned earlier during the baseline noise tests that the GPU fans have an unpleasant acoustic character but thankfully it's not as pronounced when the rest of the fans in the system are turned on. It's not exactly smooth as there is an audible low-pitched hum, but overall it sounds relatively innocuous. The system is noisier than our Rosewill Legacy W1 build by 1~2 dB, but that machine was assembled with a fanless power supply, so it may be that the Seasonic G Series 550W's fan is responsible for this difference. Still, it's very quiet even by our standards, and we'd be more than happy to use it as a gaming PC.



Previous 1 2 3 4 5 6 7 Next

Silent PC Build Guides - Article Index
Help support this site, buy from one of our affiliate retailers!
Search: