View Single Post
Old 18-03-2011, 20:33   #134
qasdfdsaq
cf.mega poster
 
Join Date: Aug 2004
Posts: 11,207
qasdfdsaq is cast in bronzeqasdfdsaq is cast in bronzeqasdfdsaq is cast in bronzeqasdfdsaq is cast in bronze
qasdfdsaq is cast in bronzeqasdfdsaq is cast in bronzeqasdfdsaq is cast in bronzeqasdfdsaq is cast in bronzeqasdfdsaq is cast in bronzeqasdfdsaq is cast in bronzeqasdfdsaq is cast in bronzeqasdfdsaq is cast in bronzeqasdfdsaq is cast in bronzeqasdfdsaq is cast in bronzeqasdfdsaq is cast in bronzeqasdfdsaq is cast in bronzeqasdfdsaq is cast in bronzeqasdfdsaq is cast in bronze
Re: Finally arrives here !

Hmm interesting. I've managed 203mbps unidirectional (UDP) at 2x2 MIMO, using standard consumer kit (DIR-825 + Intel 5100 client). Better may be possible in a more optimal configuration as there were various limitations in my setup not related to the underlying technology.

I may have mis-quoted my own number on TCP throughput though. I managed 106mps TCP on 1x1 (non-MIMO) @150 rate and 137mbps TCP on 2x1 @300 rate, but was limited by CPU on one device. Extrapolating from my other scaling tests, 250mbps UDP and 210mbps TCP should be possible with better equipment and drivers. I saw a big drop in efficiency in the 2x2, 40Mhz configuration but this was seemingly the driver/CPU limitation as neither 2x2 or 40Mhz on it's own showed any drop in efficiency at all.

Mind you, 3x3 client cards are quite widely available and 3x3 APs/routers are just coming into the consumer space now, I'd be curious to see how they perform but early reports suggest they're even more CPU limited.

Got a reference to those IEEE tests? I'm surprised mine are so different, considering.
qasdfdsaq is offline   Reply With Quote