Mayberry jumped back to 1998 when the U.S. based technology research
consortium Semiconductor Research Corp. had 34 members, when the
International Technology Roadmap on Semiconductors (ITRS) initiative was
started and when Intel first joined an IMEC research program.
drew a picture of remarkable consistency of ideas with many companies
pooling their research through a pre-competitive funnel which ended up
with essentially one manufacturing process used by multiple companies
for multiple products. Although Mayberry didn't mention it explicitly it
seems likely that the technical homogeneity has promoted the
foundry/fabless business model and consolidation on the manufacturing
side of the semiconductor industry.
Jump forward to 2013 and SRC
has 13 members. "There are fewer customers for the research and
perception issues with governments," said Mayberry. There is a
temptation to become a customer of such research collaborations but in
fact the industry as a whole needs to move to the left, to move upstream
and support a range of research on dissimilar processes, Mayberry
The industry demands constant innovation and novel
materials deployed in complex 3-D structures will likely become the
norm, he said. "Today we have even more choices than we have had in the
past – this is both good and bad for research," Mayberry concluded.
challenged from the floor that he was ignoring the economic crunch of
researching more technologies from an industry base that is experiencing
declining growth, declining margins and is generally under pressure,
Mayberry said it about making choices.
"It is a question of
funding the right cooperations. In the industry there is enough total
money available. It is like saving for your pension fund. A choice
between enjoying the expensive car today or putting something aside to
fund the future."
Click on image to enlarge.
What lies in the fog of technical
uncertainty? The end of the road for silicon development, or just an
inflection point, Mayberry asked of the ITF audience.