A fantastic voyage to the limits of semiconductor space
Why? And why now? Physical limits are coming into play, as chip manufacturing scales from 22 nanometers -- where transistors are so small that you could fit 100 million of them onto the head of a pin -- to 14, and then 10. Anything smaller is going to require new technology. As IBM puts it in a press release announcing the investment:
Silicon transistors, tiny switches that carry information on a chip, have been made smaller year after year, but they are approaching a point of physical limitation. Their increasingly small dimensions, now reaching the nanoscale, will prohibit any gains in performance due to the nature of silicon and the laws of physics.
IBM's five-year investment plan covers two research areas. First, Big Blue will study ways to manufacture silicon-based chips at 7 nanometers or below. Second, the company will expand efforts to manufacture "post-silicon" chips that don't face the same limitations as those used in today's computers. Read the full release if you're interested in the various "post-silicon" chip technologies Big Blue is exploring.
Smaller makes bigger, better
What IBM stock investors need to understand is that this isn't some blue-sky project. Tom Rosamilia, senior vice president of IBM's Systems and Technology Group, predicted that today's efforts will lead to "fundamentally different" hardware systems 10 years from now.
We should hope that he's right. IBM still depends on sales of large-scale, high-performance computer systems. In 2013, Rosamilia's Systems and Technology Group accounted for 15% of revenue, and 10% of operating profit. (Source: S&P Capital IQ.)
Mobile devices can also benefit from new chipmaking technology, especially now that slim smartphones and tablets possess much of the same processing power found in today's computers. But don't take my word for it; look at NVIDIA's (NASDAQ: NVDA) Tegra K1, which will reportedly bring the power of a desktop processor to Google's experimental Project Tango tablet.
IBM wouldn't benefit as much from advancements in mobile chipsets. At least, not directly. But that's also the beauty of Big Blue's model. The company would surely patent any breakthroughs that allow for new styles of high-performance, low-power chips, adding to a licensing portfolio that already collects more than $1 billion in royalties annually.
And even if it isn't first out of the gate with a solution to the "post-silicon" problem, IBM says in its press release that it has more than 500 patents in related technologies. That's a big statement backed by an even bigger $3 billion bet on the future of computing, one that I'm expecting will pay off for IBM stock owners.