Thermaltake Memory Cooler ReviewBosco - January 4, 2003
For the idle temperatures we let the machine idle at the desktop for 20 minutes, and for simulating load we used SiSoftware Sandra, running the memory benchmark test for 15 minutes. The CPU and System temperatures were recorded with Motherboard Monitor, while the memory modules were recorded using a temperature sensor gun.
The heat spreaders did absolutely nothing for the CPU and System temperatures (note: not a surprise), so no graphs of comparison were made for them.
Since the CPU was not the main device we stressed this time, it did not get as hot as usual, and in turn, the overall system temperature wasn't affected that much.
Memory Temperature Test Results
Quite a nice drop in load temperatures! A 7°C difference! But does a major temperature drop in memory really mean anything in terms of performance gain?
SiSoftware Sandra Memory Bandwidth Test Results
Our synthetic test shows that a cooler memory doesn't really perform that much better. In our case, it's safe to say that there is absolutely no difference!
Is there any real benefit to having a cooler memory then? In terms of the heat produced from overclocking, it's good to know that your modules will now have more room for temperature increases, in case you do choose to overclock them. It also looks good asthetically, as compared to standard "naked" modules. However, apart from that, there's not much else that I can see. Our synthetic performance gains were negligible, and I would think real-world performance gains would also be non-existant, or at least not enough to justify purchasing them based on performance gain alone. I think they would be best put to use for enthusiasts who simply wish to have the best air-cooled overclocking possible, and I'm sure this product will help them achieve that.
Just remember: Cooler memory does not imply instant performance gain out-of-the-box. The user will still need to do something to "motivate" the memory.