Intel's (Intel is an SVW sponsor] sale of its cell phone and handheld computer chip business to Marvell Technology Group for $600M, announced earlier today, represents a change in its strategy to apply the economics of the PC industry to cell phones, smart phones, and communications equipment markets.

In the mid to late 1990s, Craig Barrett, the former CEO of Intel, led efforts to expand the company's business into the fast growing cell phone and communications equipment markets. Through a series of large acquisitions totaling more than $11bn Intel built one of the world's largest communications chip businesses to become the leading provider of Internet infrastructure components.

Mr Barrett and his team saw that PC markets were maturing, and that server markets would become part of a global interconnected communications infrastructure. Intel's strategy was to extend its pole position--to providing silicon for the entire global digital communications infrastructure; and everything connected to it.

Intel wanted to provide silicon for every type of edge device, such as pocket computers, cell phones, PDAs, cameras, notebooks, PCs. And also provide the essential chips for servers and communications equipment that create the connecting foundation of the Internet.

Intel knows that investing during downturns is a wise strategy--that's what it does during the boom and bust cycles of the chip industry. And despite the dotcom crash, growth of the dotcom economy would return, and along with it the massive infrastructure investments by the telcos and others. Intel knows we are still at the beginning stages of the Internet and that the real growth is still to come.

But timing is everything in the tech industry where being too early can be the same as being wrong. Nobody would argue with Intel's long-term view of the future and its strategy, but managing a huge company on Wall Street expectations based on quarterly performance makes it doubly difficult to tune strategy to short-term market changes.

Also, Intel's comms business strategy was based on the assumption that it could apply the same economics that worked well in the PC industry. If it could produce powerful, off-the-shelf components that could be easily adapted to many different types of communications products, it would quickly win customers.

Intel expected that comms companies would ditch their custom chip efforts because they would be able to get to market faster using off-the-shelf Intel components, as the PC industry had done.

But Intel had trouble gaining traction in its comms groups. It reorganized them several times and sent in its best trouble-shooter, senior vp Sean Maloney.

For a while Mr Maloney's efforts seemed to be working and Intel was able to start publicizing some important design wins. But momentum was difficult to build, and the comms industry seemed to slow to adapt and adopt, the economics of the PC industry.

Intel thought that cell phone and comms equipment makers would quickly embrace the standard components practice that helped the PC industry grow at phenomenal rates in the 1980s and 1990s.

Microsoft, Intel's longtime partner in the PC business also expected to establish its operating systems in cell phones, PDAs and other comms gear such as set top boxes.

The barrier to adoption of standard platform was partly due to the specialized nature of comms gear but also due to the cell phone makers and comms equipment companies taking a look at the PC industry. It was easy for anyone to see that that profit margins for the PC makers were very low.

Dell, Hewlett-Packard and other leading PC makers were battling to eke out profit margins in the low single digits. Consolidation was rife in the PC industry as PC makers struggled to remain in business, and then IBM, the developer of industry standard PC platform exited the market.

Most of the value in the PC industry was captured by Intel and Microsoft. Intel's profit margins were in the 60% and above range. Microsoft's profit margins were in the 80% and above range on its PC operating systems and applications. While PC makers competed for slivers of the pie.

Could this happen to the cell phone and communications companies too? Probably, if they adopted a similar approach to the PC industry and used standard components and software.

The reason why Intel and Microsoft were not been able to "PC-ize" the comms industry is easy to understand if you looked at the PC makers as proxies for the future of the comms companies.

And that is why the industry standard PC platform that built the massive PC industry will likely be viewed by historians as an aberration that was caused by IBM's rush into the PC market in 1981, forcing it to use off-the-shelf components and operating systems. IBM didn't have the time to pursue its normal proprietary strategy of developing its own chips and operating systems.

Although the sale of the cell phone chips group is a bitter pill for Intel, it still remains deeply embedded in the comms business as a leading supplier of chips, and has key Wi-Fi and WiMAX wireless technologies.

Also, CEO Paul Otellini, in his second year, has demonstrated that he can make tough decisions. Sometimes large companies will drive themselves into the ground because they cannot overcome long established layers of legacy thinking and execution.

Mr Otellini can always draw on the advice of 40 years of Intel leadership; Craig Barrett is chairman, former CEO and chairman Andy Grove is senior advisor, and Gordon Moore is chairman emeritus.

Even though its comms business has been restructured yet again, Intel has managed to return return a huge amount of value to its shareholders since 1968, it is still an incredible cash machine, and it still manages to execute on many billions of dollars in bets in building cutting edge chip manufacturing facilities--the most complex manufacturing systems of our industrial age.

Plus, Intel still retains the ability to attack client and infrastructure comms markets in many different ways because of its ability to integrate masses of IP into ever smaller slivers of silicon.

- - -

Q1 2006 Investor Fact Sheet

Marvell to Purchase Intel’s Communications and Application Processor Business For $600 Million