Over the past year, the contraction of the communications semiconductor market has led to thousands of layoffs. Though initial layoffs affected scattered individuals throughout a given organization, later layoffs excised entire product teams. For example, PMC-Sierra is rumored to have closed down at least two of its recent high-profile acquisitions: Abrizio and Extreme Packet Devices.
Although there have been many reports of layoffs and division closures, there has been precious little analysis of the cuts' strategic implications. I'll attempt to summarize a few of the more significant consequences here.
A disturbing-albeit predictable-pattern is emerging among communications semiconductor vendors' business strategies. Specifically, vendors are streamlining their product portfolios and development activities to address the near-term needs of the largest system OEMs.
Extending the PMC-Sierra example, the company stated during its last quarterly financial conference call that it will focus for now on standard physical layer and Layer 2 chips. From a myopic point of view, this is all fine. But when all companies adopt the same strategy at the same time (as they are doing), there are some serious consequences.
The first consequence is that average selling price pressure is bound to get worse since all companies are targeting the same opportunities. Most venture capitalists I talk to will not touch 40-Gbit/second physical-layer companies with a 10-foot pole for precisely that reason. This is astonishing so early in the OC-768 development cycle.
The second consequence is that emerging system OEMs are in for a rocky ride. As semiconductor vendors prioritize which customers get precious application engineering support, startup system vendors are being pushed to the back of the queue. So, not only will they not get the fancy network processor or switch fabric they were counting on, but they also will not get the attention they need for lower-layer protocol devices. That's good news for the established system OEMs, since it is going to be a lot harder for the next Juniper to emerge.
The third consequence is that established vendors leave themselves vulnerable to semiconductor startups' targeting higher-layer products. This is the weakest consequence of the three discussed, since established vendors will eventually start acquiring again. So, the cycle continues.
Jeremey Donovan (jeremey.donovan@gartner.com) is a Principal Analyst at Gartner Dataquest.
RSS
RSS| 
Digital| 
Mobile
