Today IBM announced that they are swallowing Blade Network Technologies, a manufacturer of Ethernet switches for the blade enclosure and top of rack markets. Terms weren't announced, but the purchase has fueled the rumor mills that have been in high gear since the HP-Dell bidding war over 3Par.
Blade is a little-known Nortel spinoff that's been a leader in some of the technologies that are now making their way into the data center network, including data center bridging and virtual machine tracking. Since they made most of their money making fast switches for HP and IBM blade enclosures, and HP is likely to be using their 3Com group to fill that niche in the future, it made sense for Blade to cozy up to IBM rather than fight it out with Brocade, Juniper (who owned a piece of Blade), Force 10, Extreme, Arista and Enterasys for the third place position in data center switching. Blade's switches shipped in half the IBM blade centers, and Dell'Oro had them as number two in 10Gbps data center ports shipped but as the lack of any core switching tech would have left them as a niche player at best.
On the rumor front I'm hearing that Oracle will buy NetApp, as well as the old chestnut that Cisco is buying EMC. Larry Ellison's actually saying in public that he'd like the 60 percent of NetApp's business that he attributes to Oracle database loads fuels the fires. NetApp--while they continue to post great numbers as Cisco UCS and other customers figure out that NFS is a great way to host VMware storage--is too big to be small and too small to be big.
Then there are the fantasies starting about anyone buying Brocade. The financial press is rife with rumors that IBM, EMC, Oracle or Dell are thinking of buying the storage and data networking outfit. I think Brocade would give just about whoever buys them indigestion. While Brocade is the dominant player in the Fibre Channel switch market, the vast majority of that product, somewhere between 70-80 percent, is sold by the storage array vendors on an OEM basis. Depending on OEM sales means you need to maintain relations with those OEMs. That fact is one of the reasons EMC sold off their 90 percent share of McData in 2001.
If IBM, EMC or Hitachi were to buy Brocade, we can assume the other players would start emphasizing FCoE or in the case of EMC Cisco Fibre Channel switches. Since Brocade has a $2.8 billion market cap, any buyer already in the storage business would be betting at least $3.5 billion that they can get their competitors to play nice. Yes, Brocade has a strong Ethernet story to tell, and I've been a happy user of Foundry hardware, but an acquisition would likely cost a bunch of current cash flow on the Fibre Channel side.