Originally Posted by JeffDM
They shouldn't be doing that because that means each bit of data is being sent twice through the wireless network. When you are trying to find the max throughput of a wireless network, you want to test single-hop over that network with nothing else interfering or on that network.
Good points. Maybe this is factored into the review. Because if we look at no.2 for 802.11g encrypted, that's showing 27.2Mbit/sec.
If this is hop 1 Mac1 to AirportExtreme + hop 2 AirportExtreme to Mac2, this means 27.2Mbit transfered in 0.5sec from Mac1 to AirportExtreme and 27.2Mbit transfered in 0.5sec from AirportExtreme to Mac2.
This would correlate to a throughput of 54.6mbit/sec transferred for hop 1 Mac1 to AirportExtreme. Which is 100% of 802.11g rated...(!!) I know companies like Dlink have G+ "speedbooster" which can achieve this with 802.11G-only clients, maybe AirportExtreme has some "speedbooster" features for 802.11g.
This would also correlate to a throughput of almost 100mbit/sec transferred for hop 1 Mac1 to AirportExtreme for the 802.11n-only@5ghz mode.
We'll need better data and testing benchmarks of draft-N devices including Airport Extreme. Data tested should also be virtually uncompressible information like mp3 files. And we need to know whether their figures are 1 hop or 2 hops.
Whichever way we interpret the data in this review, it shows only as many have mentioned, a 2x increase in speed over 802.11g, not 5x. But that extra speed *is* nice.