I've not tested with AC, but using N 5GHz, I've found you had to be within a few meters, with no-one else around and any intervening structure to get the best rates. In the 5GHz waveband (as used by AC) the link rate can step down to lower levels surprisingly quickly. The kit looks well matched - both claim support for three streams at up to MCS 7 which should yield a max link rate of 975mbps in ideal signalling conditions. Such tests will do a much better job of 'stressing' your local links to see what they are capable of than an Internet speed test. If you have a second device such as a PC/Mac that can be cabled to you router, I'd prefer to test using tools such as NetIO or iPerf (both free) which essentially do the same thing as an Internet 'speed test' except you host the server locally instead of out on the Internet. Personally, I'd prefer not to test using anything involving your Internet link. It would also be useful to know (roughly) what distance there is between your router & NIC, whether there is direct (unobstructed) line of sight and whether you are indoors or out (I'd presume the former.) If you cite the exact make/model of your router and NIC, we can check out the published specs.

Incidentally, Wiki's article on 802.11AC has a rather good table of (some of) the various factors such as coding schemes, number of streams, coding rates (FEC overhead effectively) that effect link rates (speeds) if you're into the numbers game.ģMBytes/s is (very roughly) 24mbps and you could do better than that using A & G, let alone N & AC so something is clearly awry.

To get the best performance out of Wi-Fi, you need the best signalling conditions in order for the protocols to be able to use all the 'tricks' in their bag to achieve the performance.

Asus pce ac68 connecting as n