I have been tweaking lately, so I have come to a point where my old database of heatsink temperatures is no longer an effective measuring tool. So I'll have to once again start over with some fresh figures. I will still not accept anything that breaks 45 degrees Celsius, though. I have long been an advocate of cool systems, so I try not to recommend anything over that 45C mark. This is going to be a bit more difficult, however, since I raised the voltage a bit to 1.8v. I was running default voltage before, so this will help to heat things up just a bit. Here's what the test system looks like:
Antec PLUS1080 Tower Case
EPoX 8KHA+ Motherboard
AMD Athlon XP 1800+ Processor
512MB Crucial PC2100 DDR Memory
X-Micro Impact Ti4200
Seagate Barracuda 40GB Hard Drive
Arctic Silver II
As stated above, the voltage was raised a notch to 1.8v while the DDR voltage is set to 2.6v. Ambient temperature was a steady 21C so all testing will be on even footing where this is concerned. Testing will continue with our tradition of a small battery of tests that do a fine job of heating up the processor and the system as well. This forces the heatsink to work hard to maintain an acceptable range of temperatures. The temperatures will be taken at idle, after a hearty fragfest in Quake III Arena, and after a continuous looping of the 3DMark2001SE Demo.
Now to make things even more interesting, I'll take two sets of temperature readings during my testing. It's good to see how well a heatsink can keep the processor cool under normal operating conditions, but what happens when this cooler gets in the hands of a person who enjoys a little extra power? To see how it handles the rigors of overclocking, I'll measure results at 1533MHz speed and then again at 1668MHz. I left the multiplier alone for the overclocked speed since I was wanting to stress out the heatsink as much as possible. By overclocking by FSB I am forcing the processor to put out more heat than by adjusting the multiplier. For this last series of tests, I adjusted the FSB to 145MHz.
Now that we have the basics figured out, lets see how well this "budget" cooler can perform.
Results - Idle
While not the lowest temperatures that I have ever recorded, the results were quite acceptable. Also consider that I have raised the core voltage slightly and recently upgraded to a GeForce4 Ti4200 video card and you can see that the case temperatures will be rising a good bit as well. There was also only a small 1.5C rise in temperatures between the default speed and the overclocked speed. This shows some promise!
Results - Quake III Arena Deathmatch
Quake III Arena has been one of the best ways to stress out a processor (or any other component for that matter). The fact that there is only a 4C rise in temperature from idle makes a statement that this economical heatsink has some guts. I've had much higher priced coolers in the past that couldn't accomplish this.
Results - 3DMark2001 Demo Loop
This one was even better! The overclocked speed only managed to raise the temperature of the processor core a measly 3C. Since we're not using the 43-CFM Delta here, this is showing that the heatsink design is paying off. Remember those linked fins from earlier? It is quite possible that the straighter fins are doing a very effective job of forcing the air in a more direct line to the base. This will in turn help keep the processor cooler since there is more airflow to dissipate the heat.