Five years ago, network equipment vendor Brocade decided it was time to diversify. The company was, and still is, the clear marketshare leader for storage networking devices, but in 2008 it branched out into IP networking with its $3 billion acquisition of Ethernet switch maker Foundry Networks.
It was a bold move that placed Brocade in a market that is entirely dominated by Cisco. “Some people thought I was a little loose in the saddle,” recalls CEO Michael Klayko.
However, Klayko argues that the advent of virtualisation has opened up an opportunity for Brocade to stake a claim in the IP networking market. Virtualisation demands a new network “topology” in the data centre, he says.
“For the last 20 years, networks have been designed with the same topology, where most of the traffic moves north to south – from the core to the edge,” he explains. “In a virtualised environment, you want to be able to move virtual machines around the data centre. That means traffic goes east to west, from server to server.”
Growing demand for IP switching technology that could support this ‘east to west’ topology was one reason why Brocade bought in to Foundry Networks (the other being the increasing use of Ethernet switches in storage networks). “We started to build our own Ethernet product lines, but it was more challenging than anticipated, so we bought Foundry.”
Since the acquisition, Brocade has developed an ‘Ethernet fabric’ technology called “virtual cluster switching” (VCS). This is based on the quirkily titled networking protocol ‘transparent interconnection of lots of links’, or TRILL for short, that is better suited to horizontal, server-to-server IP traffic than the conventional Spanning Tree networking protocol.
Launched in December 2010, VCS is embedded into Brocade’s VDX range of data centre Ethernet switches, and the company claims that most of its 400-odd VDX customers make use of it.
Brocade has a lot riding on its Ethernet fabric business. “We’re making a huge bet on this category,” he says. “We even built our own chips.”
But Klayko says the opportunity is there for the taking. “It’s a goldrush,” he says. “It’s like the storage area network 15 years ago. The first company to get out there and get customers ended up winning – and in that case it was us.”
He claims that in this Ethernet fabric “goldrush”, Brocade has a headstart on its rivals. “I’m probably 18 months ahead of everybody in the industry,” says Klayko. “Juniper has something called qFabric, which they’ve talked about for three years and I can’t find a customer; Cisco has a product called FabricPath which has not been pervasively implemented yet.
“My goal is to have between 1,000 and 1,500 customers by the time they get their Ethernet fabric products to market,” he adds.
Analyst company Gartner describes VCS as a “well thought-out solution”, and says Brocade is “well positioned for next generation, virtualised data centres.”
So far, however, Brocade has yet to set alight the Ethernet networking sector. Its Ethernet business grew by just 1% year-on-year in its latest financial quarter, to $143 million. However, storage product revenues grew 17% during the period, pushing overall revenues up 3% to $561 million, beating analysts’ estimates.
Brocade is currently in a curious position financially. It has reportedly been looking to be acquired itself since 2009. So far, however, Brocade and rumoured suitors including PC maker Dell have failed to agree a price.
When its latest figures came out, Brocade’s share price rose. Blackstone Group, the world’s largest private equity firm, which had expressed interested in a buyout, pulled out of talks when Brocade’s market capitalisation grew.
If the Ethernet fabric “goldrush” proves as lucrative as Klayko hopes, and Brocade succeeds in mining it, the company will have to wait for an acquirer with very deep pockets indeed.