Now Hear This!

Will Moore's Law soon become no more? A new report from iSuppli Corp suggests that the law, named after Intel co-founder Gordon Moore and making up much of the foundation of the semiconductor industry, could be academic by 2014.

ISuppli argues that the high cost of semiconductor manufacturing equipment is making continued chip-making advancements too expensive for volume production. That, in turn, relegates Moore’s Law to the laboratory and "alters the fundamental economics of the industry," according to the market research company.

“The usable limit for semiconductor process technology will be reached when chip process geometries shrink to be smaller than 20 nm, to 18-nm nodes,” said Len Jelinek, director and chief analyst, semiconductor manufacturing, for iSuppli, in a statement. “At those nodes, the industry will start getting to the point where semiconductor manufacturing tools are too expensive to depreciate with volume production, ie, their costs will be so high, that the value of their lifetime productivity can never justify it.” 

Noting that while further advances in shrinking process geometries can be achieved after the 20-nm to 18-nm nodes, iSuppli estimated that Moore’s Law will no longer drive volume semiconductor production come 2014.
 
This argument isn't a new one. We've heard that Moore's Law was fading before, with the argument made in various papers, presentations, and event keynotes. In fact, even Moore himself has stated that the law cannot continue forever. (Read this 2007 Electronic News interview with Moore.)

And to be true, many factors have changed since Moore published the paper in 1965, including economic factors. It's not big news that equipment prices are up and capital spending is down. While Gartner reported this week that the situation for equipment makers is improving and that it expects growth in 2010, 2011, and 2012, that growth won't put capital spending above 2008 levels until 2012. Moreover, 2013 capital spending is estimated to fall below 2008 levels again. 

As iSuppli noted, manufacturers have historically moved to more advanced smaller-geometry processes at a fast pace to reduce manufacturing costs, but that has caused revenue generated by a specific semiconductor geometry to fall off rapidly after it had peaked. With the rising costs of new manufacturing equipment, iSuppli pointed out that semiconductor processes are expected to have more lengthy periods of revenue generation.

“The semiconductor industry will be living with historical generations of technology longer than it did before,” Jelinek said. “You are not seeing these geometries rise and fall off the way they did before. Rather, they are living on.”

As illustrated in the the below iSuppli figure, semiconductor revenue generated the 90-nm micron geometry dropped after peaking. The 65-nm geometry is not declining so dramatically after reaching its zenith, the market research company said, noting that the following geometry is likely to remain a major revenue generator for many years.
 
“Historically, the focus in the semiconductor industry was always how quickly you could move to the next geometry node,” Jelinek said. “Now the question is how to make money by sustaining a specific node.”

What are your thoughts on Moore's Law? Is it time to retire the law to academia? Have we just about reached the ultimate limit to Moore's Law, and, if so, what's next? Voice your opinion below the figure.