Arctic Cooling Super Silent 4 Ultra TC heatsink/fan

TESTING

Test Platform

The test platform is identical to the one used in the recent Zalman 7700 review.

• Intel P4-2.8A The Thermal Design Power of this P4-2.8 (533 MHz bus) is 68.4 or 69.7W depending on the version. As the CPU is a demo model without normal markings, it's not clear which version it is, so we'll round the number off to ~69W. The Maximum Power, as calculated by CPUHeat & CPUMSR, is 79W.
• AOpen AX4GE Max motherboard - Intel 845GE Chipset; built-in VGA. The on-die CPU thermal diode monitoring system reads 2°C too high, so all readings are compensated up by this amount.
• OCZ DDRAM PC-3700, 512 MB
• Seagate Barracuda IV 40G 1-platter drive (in Smart Drive from Silicon Acoustics)
• Enermax UC-A8FATR4 multifunction monitor/fan controller w/ thermal sensors
• Seasonic Super Tornado 300 (Rev. A1)
• Arctic Silver Ceramique Thermal Compound
• Two-level plywood platform with foam damping feet. Motherboard on top; most other components below. Eases heatsink changes and setup.

Measurement & Analysis Tools

• CPUBurn processor stress software
• Motherboard Monitor 5 software to show CPU temperature
• T.H.E. KP-6M reference omni microphone
• M-Audio Tampa mic preamp
• M-Audio Firewire 410 external digital sound interface
• B&K 2203 sound level meter
  

The ambient temperature in the test lab was 19°C. Ambient noise in the lab was ~16 dBA. Maximum load temperatures were recorded >20 minutes into a CPU stress test with CPUBurn once the core temperature had stabilized.

Because the Ultra TC is sold as an integrated fan/heatsink unit, we did not follow our standard testing methodology, which calls for the use of one of the SPCR reference fans with sound levels measured at several standard fan voltages. Instead, we have tested it as it would be used out of the box: Unmodified, with the fan controlled by the thermistor.

RESULTS

First, a disclaimer: All measurements in this review have a larger margin of error than is usual for SPCR reviews. This is due to a number of factors, such as lower than usual ambient temperature and the fact that the fan was not substantially louder than the ambient noise floor during much of the testing. Other factors will become apparent later on. Furthermore, the fan speed was never fast enough to produce RPM measurements; Motherboard Monitor showed 0 RPM for the duration of the test.

Now for the good news. Your eyes do not deceive you; with our P4-2.8 Northwood at full load, the Ultra TC measures a paltry 22 dBA at one meter. At idle, the fan was spinning slowly enough that individual blades could be seen — too quietly to be measured reliably at a distance of one meter. We would estimate that the noise level at idle no more than 18-19 dBA/1m. At this level, the heatsink was competing with the Smart Drive enclosed Seagate Barracuda IV HDD for loudest component on our test bench.

The fan exhibits a faint clicking noise evident at within six inches. In practice, the low noise level from this fan would usually disappear inside a case. The quality of the noise is not as smooth as the best fans we've run across, such as the Nexus models in general, or a good Panaflo 80L sample, but the level is low enough that the difference is not particularly significant unless all your other components are at or below these measured levels.

To judge the fan noise for yourself, please download and listen to the MP3 files below. Note that the comparatives represent much more expensive and usually more complex cooling solutions.

MP3 file: Arctic Cooling Super Silent 4 Ultra TC MP3, 22 dBA/1m

Comparables:

MP3: Zalman 7000 - 5V - 22 dBA/1m

MP3: Panaflo 80L - 7V - 17 dBA/1m -- on most any heatsink

MP3: Coolermaster Hyper 48 - 9V - 21 dBA/1m

And now... for another caveat. Because the Ultra TC is thermally controlled, our results are susceptible to a flaw in our testing methodology: The use of an open bench system. While an open bench system is convenient because it allows the heatsink to be easily swapped, it does not provide an environment comparable to a typical computer case, where ambient temperature is typically 30°C or more.

To counteract this flaw, we ran another test: We continued running CPUBurn and covered the heatsink with an upended plastic box. This increased the ambient temperature above the heatsink to 35°C within about 5~10 minutes. Closer to the CPU, the tmperature was actually over 40°C, but we took the lower temp as the ambient because it seemed to be the fan intake air temp. Keep mind mind that this simulation is actually a tougher environment than a typical quiet case: The latter usually has at least one case fan and some additional airflow across the CPU area; this upended plastic box environment features no such benefit.

Our "case simulation" on an open bench system.

Increasing the ambient temperature around the heatsink had two noticeable effects.

• Our motherboard finally started picking up the RPM reading for the fan, starting at ~1150 RPM. The reliability of this number is unknown; there is likely to be a considerable margin of error at the boundary of measurable signal. It might be useful to point out that a motherboard with an automatic fan failure detection will have difficulties with this fan. Arctic Cooling recognizes this difficulty and recommends that such a feature be turned off.

• The other major effect was that while the ambient temperature rose by some 16°C, the maximum CPU temperature obtained under CPUBurn rose by only 7°C. This shows that the relationship between ambient and core temperatures is not linear because of the varying speed of the thermally controlled fan. This HSF actually performs better under hotter conditions because the fan only speeds up significantly at very high temperature. It should actually show somewhat better °C/W yet with a higher wattage CPU, because the fan should speed up some more, although the resulting temperature will probably not be any lower than obtained here.

In terms of noise, a small increase in core temperature translates into a small increase in noise, around 2 dBA/1m. This means that this heatsink would most likely be close to inaudible under ordinary use, even when installed in a typically warm case. At this noise level, a hard drive is likely to produce more noise than the Ultra TC.

Of course, the low noise level comes at a price. 65°C is quite warm for a CPU, warmer than many people feel comfortable with. While our test bench showed no signs of instability during the couple hours that we were running CPUBurn, it is conceivable that a faster, warmer Pentium 4 Prescott would not be adequately cooled by the Ultra TC. Although Arctic Cooling claims that this heatsink can be used with processors up to 3.6 GHz, such a processor run at full tilt for extended periods could go into thermal throttling with an attendant drop in performance. On the other hand, the fan might push up to the highest rated speeds in response to the increased heat and thus keep the temps down. But it is not likely to be so quiet at maximum speed.

CONCLUSIONS

Arctic Cooling has produced a good revision of their Super Silent 4 TC. 24 dBA/1m max at very high load right out of the box is excellent for a modestly priced heatsink. Cooling power is sacrificed in order to achieve this level of noise, but it is just this sacrifice that is often overlooked in heatsink development. As many SPCR regulars know, there is simply no need to maintain a low CPU temperature in an everyday system. While it is true that running a CPU at a high temperature can shorten its life, a CPU is likely to become obsolete long before it burns out even in a system running in the 50-60°C range. (Editor's Note: , I've never had even a hint of damage caused by overheating in over four years of running many CPUs routinely with temps in that range.)

The Ultra TC might have some difficulty cooling a high-powered Prescott processor. That said, our test system did not produce enough heat to push the fan anywhere near its maximum rated speed of 2300 RPM, so the performance of this heatsink in a high-power situation remains somewhat unknown. To be fair, Arctic Cooling recommends on its product page that ambient temperature not exceed 38°C and also notes that Intel's thermal throttling will help prevent CPU damage once the temperature reaches a dangerous point. On another practical level, it's difficult to imagine someone spending the big bucks on a fast P4 and then limiting himself to a $20 heatsink.

All in all, the Super Silencer 4 Ultra TC cools adequately, but not spectacularly. There is little reason to prefer it over Intel's stock heatsink on the basis of cooling power alone. However, in terms of getting the job done quietly, the Ultra TC is a gem. There are few other heatsinks that provide this noise level consistently and immediately on quick and simple installation. The top Zalman and Thermalright heatsinks — both popular choices among SPCR readers — have greater cooling potential with lower noise. However, the Ultra TC's noise level out of the box is not much higher when used with a mid or low speed P4. And there's no way they can compete in terms of price. For $20, it is hard to think of a heatsink that is better value, especially from a silencing perspective. The Ultra TC is cost-effective, target-focused engineering at its best.

Pros * Quiet out of the box * Remains quiet even under high temperatures * Plenty of headroom for the fan to increase in speed * Straightforward mounting system * Safe to transport while installed * Well priced—only US$20 * Reasonable size

Cons * Might have difficulty cooling high powered Prescotts * Little overclocking potential * Overly reliant on Intel's thermal throttling technology

Much thanks to HD Audio Visual, Australia for the Super Silent 4 Ultra TC sample.

* * *

Comment on this article in our Forums.

Help support this site, buy from one of our affiliate retailers!

Search: All ProductsComputersPhotographyElectronicsSoftwareVideo GamesMoviesMusicBooksToysOfficeJewelry & WatchesApparelHealth & BeautyHome & GardenBabies & KidsFlowers & GourmetSporting GoodsMusical Instruments