Last Updated Jan 7, 2009 12:10 PM EST
The rational that Alex Handy offered was:
It would seem that Google is fed up with [Juniper Networks's] JUNOS and all the various difficulties it can present in an environment where, frankly, no one has figured out the best practices. In a world where traffic is likely doubling every six months, it's hard to think of any type of hardware architecture that could possibly sustain itself. It would seem that Juniper hasn't figured this out either, and Google has taken it upon itself to chase down a solution of its own.Why not just work with Cisco? What software or services company builds its own router? Ah, now we're getting to the interesting part of the discussion. With all the focus on Internet ads, it's easy to forget that Google is already a hardware vendor. And then there is Android, which ties the company even more thoroughly to silicon.
Many hardware companies such as HP and IBM have, of necessity, entered the software world, and Microsoft does play in hardware. Just a few months ago, Oracle announced that it had partnered with HP to create optimized servers.
Google has a big reason to expand its horizons: Its revenue is entirely too dependent on advertising. In an economic storm, as companies pull back on their spending, that could become painful. Additional diversification might be a step welcome to both management and investors. [UPDATE: When you look at Google's revenue breakout, diversification seems urgent and not just advisable.]
We're still at the rumor stage, but it doesn't hurt to remember that Google's browser was "persistent speculation" even back in 2005. How many years does it take to bake an industrial router? That could be plenty of time for Cisco and others to get nervous, if for no other reason than Android is showing that Google's ability to expand and morph is formidable.
Mechanical router image via Flickr user Joi, CC 2.0.