For testing of the Alpha P3125, I installed the unit on a P3-450 running at 580Mhz using the supplied heatsink compound. Installation is pretty straight forward. It does take longer than the average heatsink to install because the unit comes unassembled. The first thing you have to do is attach the fans on to the heatsink cover. Alpha says the fans should be blowing out. However I notice a bit better performance with the fans blowing in.

All screws and connectors are supplied with the P125. Once the fans are on the cover, you attach the cover on the heatsink itself with 4 screws. Then you are finally ready to attach the unit to the CPU using that great mounting hardware.

The P3-450 has a temp probe taped about 1mm from the CPU core. The temp probe sends its temp reading to a Super Cooler EC2-2E system monitor. This allowed me to run my torture test and monitor the CPU temp at a glance.

P3 with temp probe

The test consist of running Prime 95 and Unreal demo loops for 30 minutes. During that time, I noted the average and highest temperature readings. Just for comparison I also ran the same tests with the AMK SECC80 and SECC80+ coolers, AMK 2X60, Vantec P3-D5030 and a Global Win VEK 32 modified for P3 use. The test with the VEK32 and AMK 2X60 was done with and without the CPU backplate. Here are the results.

The Alpha P125 recorded the lowest temp reading among the coolers tested. The performance is simply outstanding. However, you can still improve it if you mount the unit without the backplate. There are many different ways to attach the heatsink to the P3 without a backplate.Click here to see how I did it. Here are the no backplate results:

Without the backplate, the performance of the Alpha P3125 became the first cooler to average below 30C in our test. It had no problems beating the other coolers. Hell, it can outperform them with the with the backplate on!

The Alpha P3125 had no problems running the P3-450 at 580Mhz and 600Mhz. However, running at 600Mhz does require the use of a L2 cache spacer. The "blow hole" in the P3125 is not enough. The performance of this cooler is simply outstanding and with the CPU backplate removed, it's untouchable.  With a street price of $61.00, it's the most expensive cooler tested, but it's also the highest performing.

The Goods

• Highest  performing heatsink we've ever tested

• Looks damn impressive

• Copper base plate

• Great mounting hardware

• Works on Slot  1 Celerons as well

The Bads

• Fans are loud

• Takes longer to  install than most coolers

• Won't fit some motherboards or case

Our rating: 9.7/10

[Page 1] [Page 2] [Page 3] [Home]