SAN FRANCISCO -- Putting a finer point on a surprisingly bullish
semiconductor-market outlook, IC Insights said Tuesday (Sept. 18)
that fewer startups, reduced capital spending ratios and the "fab
lite" movement will reverse three decades of declining semiconductor revenue growth.
In August, IC Insights
(Scottsdale, Ariz.) President Bill McLean forecast that IC sales
growth rates will rise 54 percent to an average compound annual
growth rate (CAGR) of 8 percent between 2011-21. Since 1996, the
industry has experienced an average annual growth rate of 5.2
percent, according to the report. (At that
time, McLean looked at an historic window of 2006-11, rather
than 1996-2011, but reached the same bullish outlook).
McLean's prediction noted that unit growth rates will slow from 9.5
percent per year (1996-2011) to 7 percent in the next 10 years, but
that average selling prices will jump. The outlook runs counter to
long-held assumptions that relentless manufacturing improvements in
components and fewer "killer" high-margin apps like servers and
early communications infrastructure devices have conspired to fix IC
growth rates permanently in the low single-digits.
[Get a 10% discount on ARM TechCon 2012 conference passes by using promo code EDIT. Click here to learn about the show and register.]
McLean tied improving sales growth rates to:
Fewer startups: "The IC industry is now closed to new
major manufacturing startups. This will help moderate
over-investment in new fabs."
Fab-lite foundry movement: "This should lead to less
overspending for IC fabrication capacity."
Shrinking capital-expenditure ratios: IC Insights sees the
ratio falling from 21 percent in 2011 to 19 percent this year
and 15 pecent by the end of the decade.
Wafer size pause: 450-mm wafer manufacturing will be delayed
along with the attendant cost reduction phase for IC
@goafrit, excellent point (and the second time i've heard that in 24 hours). The ability for entrenched companies to innovate is limited by their product mindset (as Clayton Christensen has pointed out for years).
All is not doom and gloom though: I visited a company yesterday that is an interesting startup in the analog space and got funding in 2010 (Touchstone Semi). The bad news? They were the only valley IC startup to get money that year.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.