Let me begin by saying again that, yes, IMHO, the majority of enterprise-class APs shipped in 2015 will indeed be based on 802.11ac, and that we'll see the wholesale replacement of 802.11n (the decommissioning of .11n APs) beginning around 2018. 802.11ac is indeed the future of the mainstream enterprise WLAN. For the time being, though, 802.11n is far from obsolete, and I continue to encourage the purchase of (especially three-stream) .11n APs for immediate needs rather than waiting for .11ac. Remember, there are very few .11ac products on the market, especially clients, and the standard isn't even finished yet. So - no rush. But anyone assuming that .11ac will have a slow uptake needs to re-examine their analysis, and soon. And, no, I'm not assuming that 802.11ad will fill the gap; I really see a lot of promise in .11ad, but it's unlikely to go mainstream because that's now the market-designated role of .11ac. But, OK, then, on to the real controversy.
Which is: just what are the real benefits of 802.11ac, anyway? Well, they're obviously not so great that, again, I'm encouraging bandwidth-hungry organizations to delay purchases of new WLAN equipment until products based on .11ac are available (and we'll see increasing numbers of these this year). First, let's look at throughput and capacity (you can see the full table in our Tech Note; I tired everything I could to insert a table here, and I continue to be disappointed with the ease-of-use of so many Web-authoring tools today). Just in the case of 40 MHz. channels (which I think will remain common with .11ac), 802.11ac is 33% faster (comparing peak speeds for both) than .11n. This is perhaps not achievable in every case and certainly not spectacular, but nothing to sniff at, either. But given that .11ac APs will cost about the same as .11n APs, who cares? Price/performance improves regardless, and we may even see a "better n than n" effect in .11ac APs operating in .11n mode - which many will, at least in the early days and perhaps much longer. I think .11ac chipsets will very rapidly replace their .11n predecessors as the platforms for new innovations, so improved .11n performance in .11ac products is, I think, quite likely indeed.
And then there's that question of when. One analysis I read, from a very large analyst firm, stated that .11ac APs will account for just 5% of all APs in 2016, a forecast I find simply astounding. This report has only a cursory analysis as to why this might be, but I believe it's dead wrong regardless - I saw similar statements about .11n ("nobody needs that kind of throughput", "802.11g is just fine", etc.), and these clearly missed the point as well: it's about capacity, and, given similar price/performance, a very clearly emerging emphasis on .11ac from every major systems vendor (and, yes, they will all have .11ac products this year), and the other points I made above, 5% market share is simply way off the mark.
To be completely fair, one gating item in the overall success of .11ac will be the availability of Wave 2 chipsets supporting, most importantly, multi-user MIMO (MU-MIMO). But as I noted earlier, this is likely to occur sooner rather than later, further reinforcing the key point I want to make here: 802.11ac is huge and will gain significant momentum in the second half of this year. And, again, while .11n is by no means obsolete, any enterprise not currently investigating .11ac will find itself at a significant disadvantage, also again, sooner rather than later.
Mathias is a principal at Farpoint Group, a wireless advisory firm in Ashland, Mass.