Test Results: High-Output Fan

This section uses the high-output Yate Loon D12SH-12 cooling fan on each product we test, so while this section is labeled high-output it shouldn't be mistaken for highest-output. I am aware that there are much more impressive fans available, such as the 120 CFM screamers that require a bolt-on kit to retain them. But in my experience, the Yate Loon D12SH-12 is one of the best 120mm cooling fans available in regards to the noise to performance ratio. The D12SH-12 cooling fan forces an impressive 88 CFM of air at a moderately noisy 40 dbA. Personally, I can't suffer anything that produces higher sound levels than this, since gaming would then require headphones and casual computing is almost impossible.

In our last section, the OCZ Vendetta 2 proved itself worthy of high-temperature systems, even while using the stock included cooling fan. But we're moving on, because almost anyone who buys an aftermarket cooler is doing so because of the free performance waiting to be unlocked through hardware tweaking. Overclockers are already willing to take their hardware to the breaking point, so it stands to reason that reduced fan noise is not a top priority. Not very long ago Benchmark Reviews compared the top-performing coolers in our Vendetta 2 vs TRUE vs HDT-S1283 article. In that article, we tested with the high-output D12SH-12 fan and liked the results so much that each cooler tested here has received the same treatment.

The high-output testing begins at the horizontal position, similar to an HTPC or rack mount server unit. I can't realistically see anyone in their right mind using the high-output fan inside an HTPC, but desktop workstations and rack mount server units could certainly benefit here. With the high-output Yate Loon D12SH-12 fan attached to each cooler, there were seven test samples taken within our ambient temperature range. In the charts below each cooler displays a thermal difference, which is the difference between the ambient room temperature subtracted from the recorded temperature results for the processor.

At first, I had to double-check my results for a typo or a superimposed name or something, because when the Noctua NH-U12P performed at the top of our horizontal test results I was absolutely shocked. Almost like an aftershock, I then noticed that the Cooler Master Hyper Z600 nudged out the OCZ Vendetta 2. So I'm going to go into this at length, because I can already imagin the flaming that TRUE fan-boys will be doing in the enthusiast keyboard-warrior forums. To begin with the results of the top four coolers are all within the range of a single degree, and although I personally tested each product as carefully as possible there's always a small margin of error. This being said, let's compare the differences between these products.

Noctua's NH-U12P uses a bolt-through system for mounting the cooler, similar to the Xigmatek Crossbow kit we use for other coolers (like the Vendetta 2). The Thermalright Ultra-120 eXtreme also uses a bolt-through kit, as does the Cooler Master Hyper Z600... so contact pressure is extremely high for all of them. Next, the NH-U12P and Hyper Z600 both have very flat mirror finishes for the contact surface, whereas the OCZTVEND2 and TRUE have a textured and equally imperfect surfaces (although the TRUE features an uneven convex surface). All four of the top coolers have either large-gauge heat-pipes, or several pairs of rods integrated into the base. Keeping in perspective that these are the horizontal test results, and it's doubtful that many of you use an HTPC or rack-mount server chassis for your overclocked system, I will finish out the remainder of the product results before I return to our top-tier bunch.

The mid-range performance for our collection revolved around a 15°C thermal difference. Noctua's NH-C12P performs just a half-degree from the Vendetta 2, and at 15.3°C the Xigmatek HDT-S1283 performed nearly the same as its less expensive clone, the Kingwin RVT-12025. After that comes the Xigmatek Achilles S1284, followed by the SilenX IXC-120HA2 and nearly identical SunbeamTech Core Contact Freezer. Finishing up our high-output fan tests was the Xigmatek under-achiever Red Scorpion S1283, which really didn't care much at all for the horizontal orientation (remember that it dropped 4°C when turned in our stock results).

The Thermaltake V2 and DuOrb, Gigabyte G-Power 2, and GlacialTech Igloo 5750 were not included in our high-output fan tests because they used proprietary designs which prohibit reasonable fan substitutions.

Now for the test that matters most to overclockers: the high-output vertical position benchmarks. Nearly every single overclocker uses a tower computer chassis, which stands upright and orients the motherboard vertically. Additionally, it's probably a safe bet that most overclockers don't particularly mind the noise level that these high-output fans create. So for this test segment, I tested each cooler for a total of seven test samples and averaged the thermal difference results in the chart below.

When the testing was complete and the results were calculated, it surprised me to find these coolers performing so close to one another. In most cases, I would normally dismiss some portion of my results as margin of error, but each of the tests were very much identical to the entire collection for each product. For a short while I had suspected that the Thermalright Ultra-120 eXtreme would beat the entire bunch, but then I remembered how the TRUE actually did better in the horizontal position than it does in vertical environments. So it didn't really surprise me when the Ultra-120 eXtreme produced the same (exact) thermal difference for both positions, because it actually got worse results in the vertical position testing under low-output fan testing back in the Best CPU Cooler Performance - Q1 2008 article.

Armed with a high-output Yate Loon D12SH-12 cooling fan, the Noctua NH-U12P once again performed at the very top of our results. If you're surprised, just imagine how I feel! We've already reviewed the Noctua NH-U12P with low-output fans and it never impressed us, so it appears that it is very much like the Thermalright Ultra-120 eXtreme in that it works best with high-output fans. Next was the OCZ Vendetta 2, which seems to produce a very-narrow lead ahead of the Thermalright Ultra-120 eXtreme, which is why the OCZVEND2 is our most recommended cooler since it costs half as much as the others near it. Of course, both the NH-U12P and Ultra-120 eXtreme can fit an additional fan on them for a push/pull combination, further extending performance for hardcore overclockers.

There are five coolers that performed nearly the same when the Yate Loon D12SH-12 was attached, beginning with the goliath Cooler Master Hyper Z600. The Z600 can accommodate four 120mm coolers, which is obscene by today's standards, but still performed very well with only one. Noctua's NH-C12P improved it's position in the standings for a shared 4th place, giving reason to suggest that Noctua look into designing some less-noisy high-output fans. The Xigmatek HDT-S1283 performed the same as Kingwin's Revolution RVT-12025 because they are the same exact product, while the Xigmatek Achilles S1284 tags along directly behind them. At the lower-performing segment of products is the SilenX IXC-120HA2, which posts a reasonable performance result only 3°C away from the best product tested. Next was the SunbeamTech Core Contact Freezer, followed by the Red Scorpion at 4°C behind the leader.

Please continue on to the final thoughts and conclusion section to read my remarks on the test outcome.

<< Prev - Next >>