Valley Wonk: The Procket Puzzle
It's a bit sad that Procket's last chapter is getting written in the shadow of the gazillion-dollar Google IPO. That was supposed to be Procket up there, clutching the gold statuette, trying to sound humble in its acceptance speech.
Instead, Procket Networks spent itself into near oblivion and is getting glommed by Cisco for $89 million, chump change that can probably be taken out of Cisco's coffee budget.
And yet people in Silicon Valley can't stop talking about Procket. It's the Apple of routers, generating heated emotions in both directions. Every core-routing thread on the Light Reading message boards mysteriously morphs into a Procket argument. Believers sing the praises of the PRO/8800; detractors seem offended that Procket even existed.
Now it's a waiting game. Sometime before October, once the paperwork shuffle ends, Procket likely will fade out. Its intellectual property and some of its engineers will get folded into Cisco, while the routers themselves become telecom's version of 8-tracks (see Cisco to Pay $89M for Procket Assets and Procket Reaches 'End of Life').
But Procket still fascinates, partly because Cisco hasn't fully explained why it's buying the company – at least, not to anyone's satisfaction. Cisco's official comment has been that it was buying engineering talent.
That doesn't stop everybody from having their own theories, many of which abounded at Supercomm last month; plenty more are in the works. Here's a smattering of the best:
- Defense: Zone coverage. Everyone agrees that Cisco bought Procket to prevent anyone else from buying it – in particular, Foundry Networks Inc. (Nasdaq: FDRY), which reportedly was in a buying mood. Some believe, in fact, that Cisco's plan for Procket is nothing – that the deal was purely defensive and the technology is getting the Raiders of the Lost Ark treatment. The acquisition amounts to "a door stop," one source says.
- Defense: Man-to-man coverage. A tastier variation on that theory says Huawei Technologies Co. Ltd. was a serious bidder for Procket. This would pretty much force Cisco's hand, as the company wouldn't want its low-cost competitor nabbing the router equivalent of weapons of mass destruction. It's plausible; Huawei has been browsing the bargain bin for crushed startups, and Procket would have been a useful addition (see Huawei Pondering Startup Deals).
- Fixing the chips. Competitors have gleefully noted the power consumption and associated heat dissipation of the CRS-1 (see Cisco Grabs a Guinness). Maybe Procket, born of a network processor operation, holds the key. It's a question Cisco will have to face as it pushes the CRS-1 architecture towards the edge; in fact, analysts say the edge version will be more significant to Cisco's future than the CRS-1 itself (see Sources: Cisco Building 'Son of HFR').
- Fixing other stuff. Others think the Procket acquisition is more of an incremental play, meant to patch up the way the CRS-1 operates. One theory says the CRS-1 needs improvement with IPv6, for example. The problem here is that one or two "off" features don't seem as if they'd justify a Procket bid, although they might have helped nudge Cisco towards the bargaining table.
- Roland the Spy. This one started before the $89 million deal ever germinated. Some say Cisco's Roland Acra was picked as Procket's last CEO specifically to get the companies to hook up. One assumes this would have been an above-board operation: Sure, Roland could skulk around and manipulate things behind the scenes, but it wouldn't take Columbo (or even a Clouseau) to figure that one out.
- Divine intervention. And lo, the clouds did part, and the Fourth Angel of BGP did appear unto John Chambers saying, "Get thee to Procket." Granted, this one came up in the wee hours of the Light Reading Supercomm Party...
There are crazier Procket theories. Here's one based on time-tested acquisition techniques of the past:
- The patent/product sweepstakes. During the bubble, startups would angle to get acquired by Cisco. They would second-guess Cisco's roadmap, then – with or without product in hand – apply for patents vital to that roadmap. If they guessed right, they had a shot at Cisco making a quick deal.
Obviously, Procket did not do this. But what if, in Cisco's grand unveiling of the CRS-1 architecture, somebody noticed an overlap with the Procket portfolio? Procket might then suggest an acquisition to prevent some messy legal questions down the road. Or, in a more twisted chain of logic, what if Cisco saw a way to cause problems with competitors approaching on the core-router path?
For the record, Cisco officials declined to comment on any of these theories (can you believe it?). As stated above, their stance all along has been that they bought Procket for the engineering talent.
The truth is probably a mix of all these factors, and the result could be a rebirth of some Procket concepts under the Cisco banner. On the other hand, if the door-stop theory is correct, the Procket engineers might flee after getting assigned to dead-end projects, but that process could take a year or two. Either way, it appears we'll have Procket to kick around for quite a while yet.
— Craig Matsumoto, Senior Editor, Light Reading