TweakTown
Tech content trusted by users in North America and around the world
6,209 Reviews & Articles | 40,188 News Posts
TRENDING NOW: GTA V will have 'just for PC' features, including a video editor

Thermaltake SubZero4G AMD Cooling System Review - Testing

For many, the mere sound of the term Alternative Cooling is enough to make them cringe. But what if I were to tell you that someone has placed some safeguards into the mix? Come join Mike "Darthtanion" Wright as he takes a hard look at the Thermaltake SubZero4G Cooling System. It uses a peltier cooling device but with some added features. He'll be looking at normal operation of this device as well as some results for an overclocked system, so come see if this is your chance to get into the new cooling craze!

| Editorials in Cases, Cooling & PSU | Posted: May 19, 2003 4:00 am
TweakTown Rating: 6.5%Manufacturer: Thermaltake

Testing

 

Now for the moment of truth. Can the SubZero4G put out enough cooling power to handle an enthusiast rig? To begin with, it should be noted that the little brain box (you remember, that PCI card) is set up to automatically adjust the power to the peltier depending on the temperature being recorded from the thermal probe. The maximum power it is rated at looks to be 73 watts from the spec sheets provided from Thermaltake. Already there seems that there might be a small problem; namely that the high-end processors put out more heat than this when overclocked even at a moderate level. But we're here to find out facts so on we go.

 

Testing conditions will follow my standard guidelines concerning coolers. Ambient temperature is kept at a regulated 21C and the cooling system will be tested in a closed case. Before I jump into system conditions we'll take a look at the test setup.

 

Xoxide modified Lian-Li PC60 Case

 

Soltek SL-75MRN-L Motherboard

 

AMD Athlon XP 1800+ Processor (Palomino)

 

1024MB Crucial PC2700 DDR Memory

 

Sapphire Radeon 9700 Pro "Ultimate Edition"

 

Western Digital 80GB Hard Drive

 

Arctic Silver III

 

I am using an older based Palomino processor since they tend to put out a bit more heat. And after all, we want to see how it handles this very thing so it suits for testing purposes. The processor voltage was set to 1.8v and memory was set to 2.6v. Temperatures are measured at idle after 15 minutes of no activity, measured again after a rugged Quake III Arena Deathmatch, and a final time after a continuous demo looping of 3DMark2001. After these temperatures have been recorded, we'll boost the FSB to 145MHz and run it through the same series of tests. The moderate overclock of only the FSB will allow the processor to do the work and not share the load across the entire system. We're after heat and we will get it.

 

A final note: Each category will have two screenshots of the temperature results. The top set of results will be at default speeds (1533MHz and 69.8 watts) and the bottom results will be at the overclocked speeds (1668MHz and 73.8 watts). The results should speak for themselves, but I wanted to make sure there wasn't any confusion here.

 

Results - Idle

 

 

 

While the temperatures aren't totally out of the water, they certainly aren't what I had expected when talking of alternative cooling. But before we pass judgment, lets make sure to put some stress to the system and see what happens.

 

Results - Quake III Arena Deathmatch

 

 

 

Interestingly we ended up with the same maximum temperature for both the default and the overclocked speeds. When we look at the heat output, though, we see that both are getting on towards the top end of the spectrum of what this cooler is capable of handling.

 

Results - 3DMark2001 Demo Loop

 

 

 

Again we see the same result with both speeds resulting in a temperature of 42 degrees Celsius. I can say that this is the first time that I have ever had a cooler with this consistent of a temperature range.

 

A bit more extreme

 

Generally I would stop my testing at this point, but I wanted to see what happened when we went a good bit above the rated heat output of the peltier. I mean if we are going to overclock, we are very likely going to be well in excess of 73 watts anyway.

 

The victim in this final test was a little Thoroughbred 1700+ processor. But to get things heated up nicely, I overclocked it to 2004MHz and set the core voltage to 1.75v, which is 0.1v higher than default. This results in a heat output of 80.0 watts. All other system settings and ambient temperatures were consistent with previous tests.

 

To save you the effort of having to sort through yet another graph, I'll just state that the results were a bit off the mark. Idle temperatures with the SubZero4G unit installed were 41 degrees Celsius and the result after a refreshing round in a Quake III Arena Deathmatch were a dismal 47.5 degrees Celsius. To give you an idea how it rates with a reasonable air cooler, I ran these same tests using a Cooler Master X-Dream SE heatsink and had results of 35.5C and 40.5C respectively.

 

So as you can see for yourself, when you start getting well above the 73 watt limit of this peltier cooling system, you start getting very drastic changes in temperature.

 

Further Reading: Read and find more Cases, Cooling & PSU content at our Cases, Cooling & PSU reviews, guides and articles index page.

Do you get our RSS feed? Get It!

Got an opinion on this content? Post a comment below!

Latest News Posts

View More News Posts

Forum Activity

View More Forum Posts

Press Releases

View More Press Releases