RSS Feed

HCW Tech Blog

For the latest info on computer hardware, tech, news, video games, software tips, and Linux, check out our new improved front page: HCW Tech Blog

Reviewed By: Ed Lau [08.22.02]
Edited By: Carl Nelson
Manufactured By: Thermaltake

Discuss this review in our forum!


About The Testing Method

As you know, we've been recording temperatures using thermistors mounted directly on the CPU for the past months as it is more accurate than thermistors on the motherboard, which are usually mounted in the socket under the chip.

However, even this has a margin of error. New motherboards are finally allowing the use of Athlon XP's built-in diode (only the KT3 Ultra-series and the Asus A7V333 currently include this feature for AMD based solutions). This method of temperature measure should prove to be even more accurate than our previous testing method so we will be using this procedure for this review.

Due to a freakish accident on my part, the V9 actually crushed my 1800+ AXP. Blame my stupidity...but this also makes me think about the equally freakish size of the V9. Anyways, lowlight lent me a 2100+ to test with while my 1800+ is sent to AMD, w00t!

The test machine consists of the aforementioned 2100+ (1.73GHz) Athlon XP mounted on the aforementioned MSI KT3 Ultra-ARU with one stick of 512MB of Samsung PC2700 DDR333 RAM. The case is an InWin Q500N with 3 case fans.

First, I will do testing with the thermo-resistor setting. My rig will be torture tested with Prime95 for 25 minutes and then disengaged. It will be left standing for 15 minutes and then the case side will be taken off. Secondly, I will use the manual fan speed wheel, running the Smart Fan II at the lowest and highest settings.

Ambient idle case temperature was 34 Celsius...yes, it's damn hot here in Vancouver. 28.5 degrees outside.

Temp Sensor Mode Test Results

As you can see, the V9's smart fan follows the temperature of the CPU very closely.

While this is a very cool feature, the V9's actual performance is lackluster. It is a very good cooler but it isn't - for lack of a better word - hardcore enough. The smart fan spins just fast enough to maintain a stable temperature. What this means is that you won't any large temperature spikes. Instead, the V9 cools very consistently, maintaining 51 degrees during the torture test, only 7 degrees higher than idle, taking a good 10 minutes to get there. With the Volcano 7, we saw a temperature difference of approximately 13 degrees.

This is all great since we know the V9 can keep an airflow well enough to not allow the CPU to heat up past 51 degrees, but on the downside, it also takes forever to cool down. Because it follows the CPU temperature so closely, the fan slows as the CPU cools instead of keeping a high speed until the desired low idle temperature. Even after Prime95 was disengaged, the CPU temperature only cooled by 4 degrees after 15 minutes and took almost another 15 after the case side was removed to return to the original idle temperature. Pretty weak.

Temperature Control Mode Test Results

Personally, I think using the manual fan control is a better choice. When you're not doing anything with the computer or if you leave it on 24/7 in your room like I do, you can turn it down to the lowest setting (which is whisper quiet) so you can sleep easier. When you're gaming, you can turn it all the way up as the sound of the fan (which isn't as bad as the 7000rpm Deltas I was using on the 102A) is drowned out by the screams of the slain or the roar of a Porsche.

Low RPM (1713rpm)

Idle - 53 C
Load - N/A

For this, I left the computer running for an hour with the fan at the lowest setting. After seeing it at 53 freakin' degrees…I decided that I would rather not do a stress test at this speed. Even though my mobo shuts down the whole system if the CPU gets too hot…let's just not take the chance if I don't have to. I already busted one CPU. =)

High RPM (5273 rpm)

Time needed to cool = 8 minutes.
Idle - 39 C
Load - 51 C

For this test, I turned the fan from the lowest to the highest and when it leveled off, I turned on Prime. As you can see, the load temp is the same as if I used the thermal sensor but the idle temp is significantly lower. Why? Because I maintained the highest fan speed instead of allowing it to slow down. This also accounts for the faster cooling time. A 14 degree drop in 8 minutes as opposed to the sluggish cooling with the thermal sensor mode.

Next Page: (4)