Will the third quarter of 2004 be seen as the one where wireless LAN switches started to look like a real market sector? Or the one where the market split into three, some of which will crash and burn? The answer matters to IT managers - who need to back a winner in their infrastructure choices.

The market share figures are out (from at least two places), and the story is a good one. WLAN switches are making real money and the leading player has some real competition. Symbol still has more than half the WLAN switch market but start-up Airespace claims the lion's share of the rest.

There's a big technology argument going on (see WLAN switches: it's hardware versus software). But what about the figures? Who wins the business wins the technology war - so what do the figures actually show?

Do the figures tell the story?
"The numbers tell the whole story," says Alan Cohen, vice president of marketing at Airespace, in a brief-but-triumphalist press release, claiming that "Airespace owns approximately 30 percent of the total worldwide WLAN switch/controller market," according to Dell'Oro.

He's wrong, of course, One percentage doesn't tell the whole story. For one thing, there's the definition of what a wireless LAN switch/controller is. It's only one part of the enterprise WLAN market, and both Dell'Oro and Synergy define it fairly closely.

According to Dell'Oro, the overall enterprise WLAN market in the third quarter of 2004 was about $212 million, roughly equal to the SOHO market and the NIC market (NICs get used in both settings, so they count separately). It is now the biggest sector, since SOHO access points are now dirt-cheap, according to Dell'Oro analyst Greg Collins. However, 87 percent of that market is for "fat" access points, of the kind sold by Cisco and Proxim. The rest goes mostly to wireless switches.

There are other ways to manage access points, including Cisco's WLSM module that plugs into its wired switches, and Extreme's approach of

Cisco still rules
Look at the overall market for enterprise wireless LANs, and Cisco is still king, with 43 percent, according to Synergy. Symbol and Airespace come second and third, on 15.9 percent and 5.7 percent each. For them, the headline news is just that they have pushed Proxim and 3Com, the other enterprise access point makers, into joint fourth place, both on 4.4 percent.

The figures also preserve the wireless switch segment as a separate sector, by keeping out any general purpose networking equipment that is used to manage wireless LANs, such as Cisco's recently launched WLSM module which upgrades a Catalyst switch to manage its own fat access points, and Extreme's general purpose Power-over-Ethernet switches that manage access points.

Forget appliances
And the figures reveal one other fact: security appliances used to manage wireless LANs are yesterday's news. ReefEdge, Blusocket and Vernier have been selling security gateways - essentially a firewall that separates wireless access points from the rest of the network - for longer than the WLAN start-ups, and have achieved a smaller share.

Dell'Oro puts these "appliances" in a separate category, which only makes about three percent of the enterprise WLAN market. Synergy lumps them in with the wireless switches (well, access points connect to them, so they look a bit like wireless switches topologically), but they don't benefit form the comparison.

According to Synergy, ReefEdge has 7.5 percent of the wireless switch market, while Bluesocket has 5.7 percent. Despite an OEM agreement with HP, Vernier only contributes 3.5 percent of the market. Don't count them out yet, but both Airespace and Aruba are ahead of them, from a later start.

Where is Trapeze?
These figures certainly seem to sort out which start-up is doing best. Airespace comes out ahead of Aruba - let's take Synergy figures this time, which gives them 24.5 and 12.1 percent of the wireless switch market respectively.

The other one you'd expect to find there is Trapeze - longer established than either, and with a bigger presence in Europe. It had trouble in 2003, laying off staff, but recently gained an OEM deal with 3Com (instead of 3Com making its own switch as it said it would in May). So why doesn't Trapeze register at all on either analyst's radar screen?

The answer is because Trapeze isn't talking to them yet. Giving figures to an analyst is optional, and so far Trapeze isn't talking: "There's this waiting game to when you report your numbers," says marketing director at Trapeze, Brian Van Nice. He says that next quarter the 3Com deal should start to make some money, and the company will probably start to play ball. "We're not ready to report yet, but we will."

So, should Symbol worry?
Despite Airespace's big share of the revenue, Symbol reckons it is still further ahead than the revenue figures suggest. In unit terms, it has 69 percent according to Synergy - simply because a Symbol system is cheaper per port, according to director of product marketing at Symbol, Graham Melville. In unit terms, Airespace drops back to 14 percent.

The market share figures show Symbol still has the lead, but it has to convince users, against an enticing set of product features from the competition, that it has the better product. That technology struggle is still just getting started.

What do you think?
Can Symbol fend of the start-ups, or is the rise of commodity hardware inevitable. Join our Forum and discuss it.