The Coming Intel-IBM Foundry Battle

Feb.28.12 | About: Intel Corporation (INTC)

Seen in strictly business terms Intel's (NASDAQ:INTC) decision to support LibreOffice in contrast to the Apache OpenOffice project makes no sense.

Most reporters focused on a budding rivalry with Microsoft (NASDAQ:MSFT), noting that LibreOffice will support a variety of mobile platforms and challenge Microsoft for business from customers like the VA.

That doesn't explain why IBM, which is supporting the OpenOffice code base within the Apache Foundation, isn't getting any Intel love when it calls for an end to the code fork.

What's wrong with IBM, in Intel's eyes, may have little to do with office software and a lot to do with Moore's Second Law.

Everyone loves Moore's Law. It's the heart of tech prosperity. Chips get better-and-better as circuit lines get closer together, meaning they can run faster-and-faster for less-and-less money. But we forget Moore's Second Law, which is that as these circuit lines get closer together it costs more-and-more money to start production.

Over time, this has made chip production a capital sink. As the cost of entry has risen, there are fewer competitors. Even Intel's arch-rival, AMD, has spun out its chip-making factory into a company called Global Foundries, backed by Arab capital.

And here's where the story gets interesting. IBM, seeing these trends, has decided to go into business with Global Foundries, at a plant in New York. Some have speculated IBM might one day want to get out of chip-making entirely and become, like most other chip outfits a "fabless" company that just buys manufacturing capacity as-needed.

Intel, seeing the handwriting on the wall, is also becoming a third-party foundry, in order to spread the cost of plants across more customers. Chip hardware is continuing to evolve into a software business, where the key value is in the design, where chips are tied to customers and specific products, with the actual production a closely-held commodity.

As Intel makes this necessary transition it is increasingly going to run into IBM, and Global Foundries. It will move from just accepting foundry business to actively seeking it out, and IBM (or IBM-Global) is going to be its rival in that business.

The OpenOffice split is the "canary in the coal mine" concerning this evolving reality. Intel needs all the goodwill it can get, among companies that might want it to build chips for them, especially in Europe, which is being priced out of the foundry business entirely. The LibreOffice move is cheap marketing for a much larger opportunity.

An opportunity in pursuit of which it's on a collision course with IBM. Smart investors will understand Moore's Second Law, because there's more capital at stake in this Intel-IBM collision than meets the eye.

Disclosure: I am long IBM, INTC, MSFT.