All problems aside, the Vapochill XE II is an impressive product and the review sample kept the 3.46GHz Pentium 4 Extreme Edition processor at a cool -33 degrees during all the tests. Although there were limitations to how far the test system could be pushed, close to 4GHz was not a problem. Using an MSI 925XE Neo Platinum motherboard and 1GB of PC5300 Crucial Ballistix memory I managed to push the system over 4GHz but it wasn’t quite stable enough to run a full set of benchmarks. In the end I had to settle for 3.9GHz on a quad pumped 325MHz bus by lowering the CPU multiplier to 12x rather than its default 13x.
At default speeds, with a GeForce 6800GT graphics card, the SYSmark 2004 score was 195, which rose to 211 at 3.9GHz – although that’s a mere 8.2 per cent increase, a few points at the upper end of SYSmark 2004 represent a decent performance difference. The CPU score in PCMark 2004 shows a much larger increase up from 5337 to 6029, an increase of almost 13 per cent. This is however based on CPU overclocking only as no other part of the system was overclocked. So you could squeeze a lot more performance out of your PC if you tweaked all the other parts as well.
Looking at some current games, Doom 3 gained over 14fps in 1,024 x 768 with 8x anisotropic filtering enabled. With 4x anti-aliasing this gap was a mere 0.6fps, but this is of course down to the graphics card as well. At 1,600 x 1,200 the results were pretty much identical, which goes to show that a fast CPU is not as important as it used to be when running games at high resolutions.
Far Cry gained 9.4fps at 1,024 x 768, but this dropped to 1.2fps at 1,600 x 1,200. However, Half-Life 2 didn’t show any major jumps with an improvement of only 1.3fps at 1,024 x 768 and no gain at all at 1,600 x 1,200. So, if you’re a gamer and thinking of getting a Vapochill, make sure your graphics card has enough grunt to keep up with the overclocked CPU.
The Vapochill products have always been aimed at the high-end enthusiast market and are likely to remain so for a while yet. I was disappointed by the build quality of the case and the poor instructions, but much of this could be easily rectified. Some sound proofing would also be a better investment than the clear windows on the side of the compressor, to dispel some of the fridge sound effects.
As I said in the beginning of this review, a Vapochill case is not cheap and this review unit will set you back £571.16 while the version for Socket-478 and AMD Athlon 64 processors is only slightly cheaper at £543.14. Personally I couldn’t imagine spending this much money on high-end cooling, but then again, I’m not trying to get every last MHz out of my PC. Perhaps if the case looked as good as it performed, the price might be a bit easier to swallow.
Asetek has produced one of the most impressive cooling solutions around and it will allow you squeeze every last drop of performance from your CPU. The Vapochill is far from cheap, but just like modding a sports car, you pay a high price for that extra performance.