Lessons Learned from the Infinera Experience?

Implications: 1.      Infinera decided to keep its chip as a proprietary ASIC. 2.      So, it has been the only manufacturer of the technology. 3.      What could Infinera have done differently to improve its competitive position in the market?  

Analysis:  In going down the path of totally owning its PIC technology, Infinera arguably opened itself up to being a bottleneck for its customers in terms of the supply chain.  It also provided a gate for its buyers to get new technologies.  There is no question that it has taken Infinera a long time to develop native 40G capability.  And as a public company, the odds are lower that it is going to spin new versions of its PIC very often because it would be very expensive.   

Infinera could have considered licensing its PICs to other manufacturers so that it would get some help in furthering the technology, in particular, up the yield ramp.  Other vendors would have made it easier to improve the process and also would have provided multiple sources for it on the systems side of its business.  

The path that a startup chip vendor, Ekinops, has taken in going all the way up the food chain -- is providing the value in the firmware.  The supplier wanted to make sure that it could use off-the-shelf hardware packages.  The idea was to avoid not getting tied to a technology that costs tens of millions of dollars to spend every time it wanted to introduce a new flavor.  In addition, Ekinops argues that it achieves economies of scale because it takes advantage of the volume benefits of using hardware from larger manufacturers.  

In fairness, it should be pointed out that Ekinops is unable to upgrade line-side cards in the field, such as Ethernet, from 10G to 40G.