AeroCool X-Factor CPU Cooler (HSF) Review
Reviewed by: Bosco
Reviewed on: December 11, 2002
: GF City Computers
Price: $35 USD
I recently reviewed the AeroCool AB7080H, as part of my ongoing search for a quiet but efficient cooling solution for my AMD Thunderbird 1400. Today I will interrupt my stalking of bosco on the MOH servers to look at another AeroCool heatsink/fan combination from a different series, the X-Factor.
The X-Factor is from AeroCool's Extrusion Fin Series. What does that mean you ask? Well, Websters dictionary defines "extrusion" as coming from the verb "extrude":
1: to force, press, or push out
2: to shape (as metal or plastic) by forcing through a die.
The main body and fins of the X-Factor are made of extruded aluminum, with a copper center core which contacts the processor. This is an increasingly common design configuration which makes sense from the perspective of physics - effectively, copper is very efficient at conducting heat (ie., from the cpu) and aluminum is more efficient at dissipating, or getting rid of, the heat. Let's have a look.
The box is nicely designed, with a black and silver theme. Once again there is no literature or installation instructions in the box, just the HSF and a small soft-pak of Stars Heatsink Compound. My first impression of the X-Factor was very positive, the aluminum looks great and contrasts nicely with the black of the fan. The X-Factor is also available with an aluminum fan, I haven't seen it in person but the picture looks spectacular. The aluminum fan is rated for slightly lower performance than the review unit.
Turning the unit over, there is a clear plastic sticker across the bottom, to protect the copper from marking. I believe this is a good practice on the part of AeroCool, as the previous unit we reviewed did NOT have any protection and the bottom of heatsink was fairly badly scratched in places. The sticker is clearly marked to be removed before installation, as you can see. Although the finish of the copper is very shiny, there are noticable machining marks on the copper core, additionally, as I ran my finger across the aluminum and onto the copper, I could feel a small ridge - the copper is not 100% flush to the aluminum body and sticks out a bit. Providing the processor is totally under the copper core this should not be a problem, however, if the X-Factor were to be used on a large processor or if it were mounted slightly off center, I can imagine a scenario where damage would result. My processor did fit under the copper core although the edge of the core was only a fraction of a millimeter from the edge of my processor. Although for the purposes of the review I have not lapped the heatsink, I absolutely recommend that if you purchase this unit, you get out the sandpaper and dish soap and give it a lapping.
Taking the fan off, you can see that rubber spacers are used between the fan and the body of the heatsink. Any time you can reduce potential vibrations, you are reducing both noise and long term wear and tear, so this is a great idea. Although it is strictly an aesthetic (looks) beef and not functional, it would have been nice to see the top of the heatsink and core finished to the same level as the rest of the unit. The 70mm ball bearing fan is an EVENFLOW R127025DU rated for 5400 RPM and 44.25cfm @ 39dba. Motherboard Monitor 5 readings indicate an RPM range of 5625-5818. The box calls this fan the "Silent Tornado", it's not quite as quiet as a Sunon 60mm 5500 RPM fan I had installed on my OCZ Gladiator, but it is (subjectively speaking) a hair quieter than the fan on the previously reviewed AeroCool AB7080H. It is worth noting that, although both AeroCool fans have identical specifications on Lux Design's website, they do have slightly different designations (.....DH for the AB7080H and ....DU for the X-Factor) and the X-Factor fan spins on average 600 RPM faster. I repeat and emphasize that the faster spinning fan was slightly quieter, there was a noticable high pitched undertone on the previously reviewed unit.
One last observation at this stage, I have no idea whether it is by design or just in manufacturing, but the fins are of very different thicknesses. This is plainly visible in the picture above. Let's see how installation goes; I had some issues on the other AeroCool.
The X-Factor uses a 3-point clip design on both sides. Getting this unit installed took less torque than the AB7080H, pushing down on the clip with my thumb and using a small straight-edged screwdriver to get the clip over the socket joints, it was installed in under 30 seconds. I was relieved, because after installing the previous unit I had scratches on the ceramic of my processor and believe I came close to doing serious damage. At no point during the installation of the X-Factor did I have any concerns about the amount of pressure I had to use, yet the 3 hole clip gives a good firm hold onto the socket.
At 70mm X 70mm X 57mm tall (2 1/2" X 2 1/2" X 2 1/4" for the metrically challenged) including the fan, I had no problems with space on my Asus A7M266 motherboard, the X-Factor easily cleared the capacitors near the socket and all other motherboard components. Time to lay a hurtin' on it and see how it responds.
All benchmarks were performed using Quicksilver thermal compound, rather than the supplied Stars compound. From a cold boot, I waited 20 to 30 minutes before using any function on my rig, in order to establish idle temperature. I loaded FreeCell and played....gotcha. Really, I fired up SuperPi to 32 million digits and ran that in the background while I looped 3D Mark SE 3 times. Recorded temperature. Put in eyedrops required after watching 3DMSE stutter that badly. Then it was onto Prime95 for 20 minutes, record temperature, and 15 minutes of looped Quake 3 timedemo at low quality settings for the final recorded temperature. All tests were run 3 times minimum from cold boot, and the results averaged. Ambient temperature was at all times 20 C and hardware temperatures were taken from Motherboad Monitor 5 or from the bios. Asus motherboards allegedly report temperatures higher than actual - I have not verified the temperatures using other monitoring hardware. All temperatures are measured using the same method, the results are reported accordingly.
The comparison is between the X-Factor, the AB7080H, an OCZ Gladiator with a 60mm 6800 RPM (7300 actual) Delta fan and the same OCZ with a 60mm 5500 RPM Sunon fan.
I will also note that I left my computer on overnight. It is located in my home office, and the window is always left open for ventilation. When I got up the following morning, the room was 17C and MBM5 showed the core temp to be 41C. It's obviously not statistically valid to be included in the graph above but I have never seen my processor at under 43C before, even under similar (long idle, cooler room) conditions.
I like it. Performance equal to or very slightly better than my howling OCZ/Delta combination, at much quieter sound levels. The X-Factor is more expensive than its cousin the AB7080H but I would not hesitate to spend an extra $10.00 Cdn to get this level of performance. The difference in both idle and load temperatures is only a couple of degrees but combined with a slightly quieter fan, the value is there. I would even conjecture that if I would have lapped the X-Factor, I might have gotten a better "seal" and maybe eeked out another degree or so of cooling. At $44.95, in the middle of the price range for air cooling, it may not be for the extreme overclocker but I would recommend it for everyday use and moderate overclocking. Now, where'd that pesky bosco go?