After reading a Wall Street Journal piece about how Juniper Networks could emerge as a potential acquirer of Brocade Communications, I thought about what it all might mean.
When Brocade or its agent Qatalyst Partners leaked that the company was up for grabs and that HP and Oracle were potential buyers, I thought probably just one of those companies had expressed solicited or unsolicited interest in buying Brocade. Since then, we’ve learned that HP has taken a look at Brocade, but has yet to decide whether it wishes to make an acquisition bid.
Oracle is playing its cards close to its vest, and understandably so. Much of Oracle’s data-center master plan turns on its owning and integrating the assets of Sun Microsystems. With the fate of the Sun deal undecided, potentially for as long as three months, Oracle isn’t in a position to make an acquisition predicated on owning and assimilating Sun.
Knowing what we seem to know – that HP has kicked the tires, but has not yet decided whether it wants to buy Brocade, and that Oracle is in an extended holding pattern – why has Brocade put up the for-sale sign?
This is where Juniper enters the picture. One could make a plausible case for Juniper wanting to acquire Brocade. It’s a possibility, yes; but I don’t think it’s probable.
That’s because Brocade would be an expensive proposition for Juniper to digest, with significant product overlap on the Foundry side of Brocade’s house. What’s more, Juniper has eschewed large acquisitions ever since its fraught purchase of firewall-market leader NetScreen Technologies back in 2004. That $4-billion all-stock deal was a bear for Juniper to integrate. It took a long time to align the NetScreen and Juniper executive teams, and the porting of NetScreen’s security software to Juniper’s JUNOS took longer than the vast majority of observers anticipated. Juniper found out the hard way that acquisitions can look good on paper, but making people, products, and technologies work together goes beyond the page.
One can make a case that Juniper might want Brocade’s fiber-channel SAN switching, but I think Juniper would rather roll its own or buy a smaller company than go through the headache and overhead (porting to JUNOS, etc.) that a costly, large-scale acquisition of Brocade would entail.
Besides – and here’s the kicker – Juniper is probably a seller, not a buyer, in the computer-networking industry’s consolidation market.
Call options on Juniper have gone up recently, with market movers and shakers wagering that the company’s stock will appreciate in value during the next few months. These option buyers wouldn’t be betting on an appreciation in Juniper’s share price if they believed the company were about to pull the trigger on a costly and complicated acquisition. The only explanation is that they are expecting the company to have a very strong fall quarter, with encouraging forward guidance, or that they are expecting the company to be sold at a reasonably attractive premium.
Meanwhile, HP seems to be considering a purchase of either Juniper or Brocade. Each company would offer HP ProCurve Networking something it doesn’t have today, but a Juniper acquisition would give it the high-end routing and data-center switching it needs to take the fight to Cisco. Brocade would be a less-ambitious acquisition, and substantially less costly, but HP might figure that it’s in an all-or-nothing situation in the data center. With it finally getting past its integration of professional-services behemoth EDS, HP could be ready to make another big buy.
An acquisition of Juniper wouldn’t be a snap for HP to pull off from a product-integration standpoint, but HP might have the confidence it can do it successfully after its experience assimilating EDS.
If HP moves on Juniper, where does that leave Brocade? Out in the cold and looking for shelter, I suspect. My guess is the HP and Oracle both looked at Brocade well before that intentional leak from Brocade or Qatalyst surfaced early this week. HP might be heading in another direction, perhaps toward Juniper, and Oracle isn’t prepared to move until the European Commission approves the Sun deal.
Brocade can hear the strains of consolidation-related musical chairs, and it wants to make sure it gets a comfortable seat. On its own, its prospects aren’t that good. But, if it could get IBM to reconsider its vendor-neutral, systems-integration approach to networking hardware, it might find sanctuary. (There might be somebody else that would be interested in Brocade, but I’m not seeing it yet.)
IBM resells Brocade switches and accounts for 20 percent of the latter’s revenue. IBM’s “Dynamic Infrastructure” vision for data-center convergence places power and value in a layer of management software that provides virtualization and automation, effectively relegating underlying networking hardware to little more than plumbing. The IBM approach is similar in some respects to HP’s, but the latter has chosen to own its networking infrastructure while IBM has opted for reseller and OEM relationships with the likes of Juniper and Brocade.
What happens, though, if HP takes Juniper off the table, effectively removing an IBM partner? Even though IBM envisions networking hardware as a less-valuable element of the overall data-center solution, would IBM respond by simply striking another deal to resell another vendor’s networking gear? Or does it reluctantly decide it has no recourse but to become a player in network infrastructure?
That question might become more than hypothetical before the end of this year.
Twilight in the Valley of the Nerds 
Regarding Brocade, you are looking in all the wrong places for a buyer. You have to look at who wants to get into the Data Center space. Who sees it as a natural place to grow and expand their server business. The answer is staring you in the face. This company has been looking at this space for some time. This company has the cash to do a deal. This company is looking to breakout of its space and grow and expand. BRCD is the target. The company is DELL!
In a follow-on post regarding Michael Dell’s comments during his swing through the Bay Area, I mentioned that Dell could be a buyer of data-center networking assets.
The challenge for Dell is two-fold: First, it has just begun digesting its acquisition of Perot Systems, which is extremely important to the company’s services-based push into healthcare and government vertical markets in the USA and China; second, Dell is strategically all over the map, with its fingers in a lot of half-baked pies.
As for as I can see, it doesn’t have a clearly articulated, fully formed strategy for data-center convergence in large enterprises. It will need one of those to guide any networking acquisition it might wish to pursue, irrespective of whether the deal would involve Brocade, Extreme Networks, or Arista Networks.