JMke's P4 Test Setup
CPU	Intel P4 2.4 "C"
Cooling	VapoChill XE II
Mainboard	Asus P4C800
Memory	1 * 256Mb PC3200 Mushkin Special Series
Video	ATI R9000 Passive Cooling

Default CPU speed: 2400Mhz
Maximum Overclock with air cooling: 3200Mhz - 1.7v = 800Mhz increase over default
Maximum Overclock with VapoChill XE II: 3725Mhz – 1.76v = 525Mhz increase over air cooling

JMke's A64 Test Setup
CPU	A64 3200+ CG Stepping
Cooling	VapoChill XE II
Mainboard	DFI NF3 250Gb
Memory	1 * 512Mb Mushkin PC3200 LVLII V2
Video	Chaintech FX5900XT

Default CPU speed: 2200Mhz
Maximum Overclock with air cooling: 2500Mhz – 1.65v = 300Mhz increase over default
Maximum Overclock with VapoChill XE II: 2915Mhz – 1.70v = 415Mhz increase over air cooling

---

The results obtained per CPU might not be the best around, but when comparing to the default speeds and the maximum obtainable with air cooling, the increase is quite noticeable, 1325Mhz on the P4 and 715Mhz on the A64!

The CPU temperature readouts are not too accurate as the onboard sensor can be placed anywhere under that IHS, but the results are comparable to each other when changing CPU speed and voltage. Using this CPU heat output calculator I set the A64 at different overclocks using a variety of vcore settings and graph the results as follows.

Room temperatures during testing was 23°C



At the theoretical maximum of 200Watt the Evaporator reads -22.6°C while the CPU is at a cozy 36°C, to show you the difference in CPU temp readouts I did a quick test on the P4 platform also and obtained quite different results when it comes down to CPU temp, however Evaporator temps remain nearly the same:



The P4 results temperature wise seem to be much more impressive then those on the A64 platform, however I don’t trust the onboard CPU probe of the P4 all too much, as it is located near the outer edge of the IHS (integrated heat spreader) and will always show temperatures lower then they actually are.

Now the Control Panel allows the fan speed of the 2x120mm Sunon fans to be changed so I wanted to know how it affected temperatures and noise production when going from 100% to 30% of the fan’s speed:

A decibel meter was placed ~50cm away from the closed VapoChill unit and the A64 was placed under full load at 2500Mhz – 1.6v vcore, here are the results:



To give a little bit more info on the dBA results, a change of 3 dBA is noticeable to most people while a 10dBA increase means it is twice as loud. Looking at the graph above this seems about right with how I interpreted the noise. The 50% fan settings cuts not only the fan's speeds in halve, but also the noise, the change to the 30% setting is hardly noticeable, however the difference between 75 and 100% is.

The performance of the XE II does not suffer all that much by decreasing the airflow, although the Evaporator makes a higher climb then the CPU temp. The Control Panel allows you to change the fan’s speed in 1% steps so finding a setting between 50-75% would be ideal for a performing/silent setup.

Now let’s see what a 715Mhz speed bump can do for the A64 in benchmarks and gaming ->