File talk:Speed-vs-length.svg

graph is out of date

This graph is seriously out of date. It would appear to indicate the maximum speed of UTP at 100 meters is around 25Mbps, when in fact gigabit ethernet (1Gbps at 100m) has been using UTP for many years now, and 10GBASE-T (10Gbps at 100m via UTP) has been around since 2006. Since the graph starts at 10 meters, it also should be noted the current IEEE 802.3ba spec (as of 2007) is for 40Gbps at 10m via UTP. As of 2002, DWDM technologies offered 1.76 Tbps at 600 km, 800 Gbps at 1500km, and 400Gbps at 2000km via single-mode fiber. --Pascal666 (talk) 20:56, 26 October 2008 (UTC)

Have you thought about the possibility that in higher speed applications e.g. multiple pairs of STP or UTP will be used? Are you aware of fiber communications using multiple light colors for transmission? What about the usage of optical repeaters for fibers? All those options are not at all covered by the diagram because the diagram does deal with the "raw" capabilities of a single media unit or segment in a networked environment. BTW the given limits are physical limits, so i wonder why physics did change between those results and your numbers. I know about fibers getting better by some few percents over the last 40 years of research and evolution, not much in total but at least slightly. A good explanation for this assumed improvement aspect would truly deliver valuable content for any wikipedia. So please provide them. Just tossing a few numbers into the pot wont help that much, at least not with creating a revised or updated version of this diagram. --Alexander.stohr (talk) 21:59, 26 October 2008 (UTC)

Now you have me truly confused. Data rate as rated in bps (or mbps, gbps, tbps, etc) is never a "physical limit". Data rates increase all the time as new standards are written. Bandwidth as rated in mhz may be considered a "physical limit", but the designation "UTP" for example does not speak to any particular limit. The earliest implementation of UTP I am aware of had a twist rate of about six per mile [sic]. Category 5e cable has a twist rate of about three per inch. These two variations of UTP obviously had very different properties. A bandwidth of 250mhz is normally quoted for Cat6, and 550mhz for Cat6a, but bandwidth (in mhz) is only loosely associated with data rate (in bps).

If this graph is indeed attempting to quantify data rate in bps, it will have to be updated regularly. I attempted to give some numbers above that can be used to update this graph. If this graph is intended to quantify "physical limits", the current labels do not make sense. Before trying to update the graph, perhaps it would be best to ask where the current numbers came from? Perhaps the original text version has different labels that can be applied to this graphic to make it accurate. --Pascal666 (talk) 19:43, 29 October 2008 (UTC)