HH Testing Platform and SANDRA XII Network Tests
How we configured our test systems: When configuring our test systems for this article, we set up a desktop and laptop system on the same network. The desktop was connected via its 100Mbps port whereas the laptop was connected using the Actiontec kit. For the test, the default - out of the box Actiontec encryption was employed and Windows Firewall was disabled on each system as well.
IBM Thinkpad T-43
Intel Pentium-M CPU
1.5GB DDR2 RAM
75GB Hard Drive (ATA)
|System 2: |
Intel Pentium-4 Williamette
ASUS P4C800 MB
2GB DDR2 RAM
BFG Technologies Asylum GeForce FX 5950 Ultra
The first half of our evaluation is going to consist of synthetic network bandwidth testing courtesy of SANDRA. In this test, the program requires at least two machines to be on the network, with SANDRA automatically identifying the second system as the host.
With Windows reporting that we had a successful 100Mbps connection via the Actiontec kit, the Network Bandwidth testing showed an average bandwidth between 54Mbps 802.11g and 100Mbps, and slightly, very slightly, below 802.11n. Not bad. We should also note that ratings for bandwidth are absolute peak and are not completely representative of real world results. With networking performance, there are a multitude of factors, including latency, that can affect the actual performance. Notice the Actiontec MegaPlug AV unit had latency characteristics that fell somewhere in between 802.11n and 802.11g (the latter providing a lower latency connection) performance but over 2x higher latency than standard 100Mb Ethernet. So, in terms of latency, as an alternative to 100Mb Ethernet, the Megaplug AV 200 held relatively strong. In our next test, we'll try to give you an idea of more real world results.