Thursday, July 23, 2015

Cisco Exits Cable TV Equipment Business

The historic concern U.S. cable TV operators have had is that, as a smallish industry, it would not gain support from a broad number of suppliers well placed to develop industry technology. In some ways, that remains a valid concern.


Consider what happened to the two major suppliers of set-top box technology, Jerrold and Scientific-Atlanta.


Jerrold became General Instrument, which was in turn purchased by Motorola, then briefly became part of Google, before being spun off to Arris Group itself a firm that had grown since about 2002 by buying companies in the cable TV infrastructure business.


Scientific-Atlanta, meanwhile, was sold to Cisco, which primarily wanted the set-top box business. Cisco now has sold the asset to French firm Technicolor, which now becomes a major supplier of customer premises equipment and network infrastructure  for the U.S. cable TV industry.


S-A originally was purchased for about $6.9 billion. One might argue that, over the last decade, the value of that business has plummeted by an order of magnitude.


Cisco is selling the business to the former Thomson for $600 million (or €550 million equivalent) in cash and stock.


In substantial part, the sale reflects Cisco’s move away from consumer businesses.


Cisco will continue to refocus our investments in service provider video towards cloud and software-based services businesses.


The Connected Devices business will end fiscal 2015 with revenue of approximately $1.8 billion.

Some things apparently have not changed over the last several decades. The extent and health of the industry supplier base remain important concerns for service providers. Where annual cable capital spending might be in the single digit billions, annual telco capital investment represents a market worth about $350 billion annually.

No comments:

Costs of Creating Machine Learning Models is Up Sharply

With the caveat that we must be careful about making linear extrapolations into the future, training costs of state-of-the-art AI models hav...