Intel Sells Optical Network Business
Continuing the company's massive reorganization program, part of which involves the transition of employees to other companies, Intel announced on Monday it has sold its optical networking components division to privately held San Jose-based Cortina Systems for an undisclosed amount.
Up to this point, Cortina's claim to fame has been a collaborative venture with Cisco Systems to develop the Interlaken protocol, a royalty-free interconnect technology for 10 gigabyte-per-second (Gbps) optical networks. If your front-line business just happens to be the production of a royalty-free product, you'd better be prepared to back that up with a technology that makes full use of it.
Today, Cortina may actually own a big piece of that technology, having just acquired many of Intel's media access controllers and network physical layer devices, as well as all of its error correction framers, transport framers, and T1 interface products, potentially elevating Cortina to the role of Cisco competitor as well as partner.
Unlike Intel's previous division sales to established firms Marvell and Eicon Networks, Cortina's acquisition yesterday was heavily leveraged, with six different venture partners receiving full credit in Cortina's statement. The hope here, apparently, is to create a completely new player in the hardware space, with an established product line and an interconnect protocol already under its purview.
Cortina's CEO Amir Nayyerhabibi was cited yesterday as saying, "This acquisition positions Cortina to become the leading pure-play communications semiconductor company." What that means is that these existing MAC and PHY semiconductor products should fare better in the market when managed by a company whose singular focus is network semiconductors.
But the relationship between Cisco and Cortina may change, now that the two will be competing in a product space where Cisco has a certified stake. While the two jointly developed the Interlaken protocol out of necessity, it might not be the only solution for tackling the problem of transmitting data over serialization/deserialization devices (called serdes) when traversing networks where the speed must ramp up. Interlaken adopts an approach to this problem that, while workable, breaks downward compatibility with an older serdes protocol.
If the two firms were to part company, conceivably, no one would be in a better position than Cisco to develop an alternative protocol that would render Interlaken unnecessary. So even while the new Cortina will try to position itself as a new and singular player in a burgeoning market, it may find itself sandwiched between two giants.