« CES 07: Wireless Port Replicator eliminates cables | Main | CES 07: Toshiba bringing out twin hard drive laptop »
Home networking speed tests
The testing of Panasonic's Ethernet-over-mains networking kit, the BL-PA100KT, threw up some interesting questions for us here at PCW.
In our BL-PA100KT review we reported TCP file transfer speeds because TCP is reliable and sends data in order. It is the staple diet of the networks and the internet.
UDP, on the other hand, is unreliable with packets going missing from time to time but it is a lot quicker because it doesn't perform the same checks.
UDP has been used for rudimentary tasks for years, like keeping networks awake, but it is now being used widely in IPTV and other media streaming; if you lose a little bit of data that doesn't matter too much because the picture will just lose a little clarity.
So, from a testing point of view it's interesting to provide two sets of data; TCP transfers and UDP transfers.
Using the Panasonic BL-PA100KT home networking kit in TCP testing (for example, word documents, photos or a big zip file) we achieved 25.4Mbits/sec but for UDP (video streaming) we averaged 78.3Mbits/sec.
These are phenomenal speed differences which make UDP transfers much more attractive for media streaming; at 78.3Mbits/sec this network could handle multiple high definition streams with ease.
PCW currently reports on TCP results as the datagram protocol of choice, but as streaming HD content becomes more of an issue in our lives we'll start to look at UDP too.
If you see vastly different speeds in different publications then specifying UDP or TCP becomes important.
Finally, like all homeplug-esque products, Panasonic claims its power line products are severely affected by electrical noise that is passed onto the electrical wiring in your home. Apparently noise can limit speed and reliability of their adaptors, however I found no such problems.
Energy efficient light bulbs are supposed to be a major offender so I conducted brief speed tests in the dark, with everything in my house turned off, then in the light, with everything turned on (energy-efficient lamps, TVs and all) and found no meaningful difference in speeds.
Posted by Emil Larsen on January 12, 2007 | Permalink
TrackBack
TrackBack URL for this entry:
http://www.typepad.com/t/trackback/24766/7470237
Listed below are links to weblogs that reference Home networking speed tests:
Comments
sp
homeplug-esk ==> plug-esque
severly ==> severely
Posted by: David W | 17 Jan 2007 23:33:24
