Benchmarks/Testing/Results/Performance

How do you test pumps? Especially ones that are designed for different size hoses? After much debate in the GruntvillE virtual office it was decided that the pumps would be tested on a system with ½ ID tubing and 3 blocks. Each pump would be installed, the system bled, and then run for a few days to ensure that there are no bubbles or air pockets in the loop. Temperature measurements were taken using Motherboard Monitor for the case and CPU temp, ATI Catalyst drivers for GPU temps, and a basic thermal probe for room temps.

Once the system was bled and ready we took idle readings. CPU load was generated by running 2 instances of Folding@Home for 6 hours before any readings where taken. Then we add in the GPU load by firing up 3DMark 05 to run along with Folding. We did not do any Overclocking tests, but it is safe to assume that the lower the temps on the tests we did, the better the chance of getting a good OC.

For the test bed I used the Megadeth case, which is a modified Lian Li PC-70 with the following hardware and cooling:

•  MSI 875P Neo FIS2-R Motherboard
•  Intel Pentium 4 3.0ghz CPU
•  (2) 512mb Corsair XMS3200XL Pro
•  HIS Radeon X800XT IceQ II video card
•  Danger Den Black Ice Extreme II dual 120mm radiator
•  (2) Panaflo 120 “M” fans
•  Primoflex 1/2ID Green Tubing
•  PolarFLO TT Series CPU Block
•  PolarFLO TT Series Chipset block
•  PolarFLO TT Series GPU BLock
•  Primochill Dual 5.25 Bay Res
•  Raidmax 500 watt PSU

When testing the MCP-350 I added 2” of 3/8ID tubing and a 3/8 to ½ adapter on both the inlet and outlet. The flow was setup like this Pump – Radiator – CPU. Since the CPU has two outlets one was connected to the chipset and one the GPU then the reservoir and back to the pump.

The following table shows the average reading for each temperature that was measured. The important thing to keep in mind is the Room/CPU/GPU delta. This is the measurement of how much higher then the ambient room temperature the CPU and GPU are. The first number is the delta for the CPU, second is the GPU. The lower this number the better the cooling provided by the system. All temperatures are in Degrees Celsius.

 

 

MCP-600

MCP-350

PolarFLO TT

IDLE

 

 

 

Room

24

24

24

System

33

33

28

CPU

33

33

33

GPU

33

33

33

Room/CPU/GPU

9 / 9

9 / 9

9 / 9

CPU Load

 

 

 

Room

24

22

24

System

30

31

30

CPU

35

38

33

GPU

35

33

34

Room/CPU/GPU

11 / 11

16 / 11

9 / 10

CPU/GPU Load

 

 

 

Room

24

24

24

System

33

33

30

CPU

35

40

33

GPU

42

42

41

Room/CPU/GPU

11 / 18

16 / 18

9 / 17

 

There are no real surprises here. The PolarFLO TT, which is based upon the Laing D4 12v pump, performs better in this particular setup then the MCP-350 while the MCP-600 just about keeps up. When using the results from this review to help you determine which pump to use please keep in mind that every water cooling system is different. Unless you plan on using the same blocks and radiator as used here your results will be different. However, if you have picked out the blocks, radiator, etc, then the differences in the pump performance can be used as a guide.

 
Next >>>>
CONTENTS
Page 1: Introduction
Page 2: Specifications & Packaging
Page 3: Installation
Page 4: Benchmarks/Testing/Results/Performance
Page 5: Conclusion


GruntvillE.com is © Copyright 2002 - 2005 GruntvillE.com. All Rights Reserved.