IEEE Solid-State Circuits Magazine - Spring 2014 - 78

ISSCC 2014 CEO and VC Panel (from left): Ali Keshavarzi (moderator), Andy Rappaport, T.J. Rodgers, John Doerr, Scott McGregor, Dado Banatao, and Nicky Lu.

Nicky Lu, founder and CEO of Etron;
Andy Rappaport, partner emeritus
at August Capital; John Doerr, managing partner at KPCB; and Dado
Banatao, founder and managing
partner at Tallwood Venture Capital.

Where Have the VC-Funded
-Semiconductor Startups Gone?
Andy Rappaport opened the discussion with a VC perspective on semiconductor investment. The number
of Series A semiconductor ventures is
decreasing-from a peak of 64 companies in 2003, to three companies in
2011, and finally to the point in 2013
where he said people no longer bother
to measure. Nicky Lu presented data
from the most-recent Global Semiconductor Alliance report showing that
funding for Series A and later financing of semiconductor companies
dropped from US$3.9 billion in 2006
to US$0.4 billion in 2013.
The costs of developing an SoC in
an advanced deep submicron technology are so prohibitive that only a
few major industry players can afford
to design chips. Most chips designed
today contain over 90% IP reuse. So
where is the innovation in these mega
SoCs? TSMC's volume is dominated by

Andy Rappaport, of August Capital (left), and
T.J. Rodgers, Cypress Semiconductor (right).

s p r i n g 2 0 14

the few customers (including Qualcomm, Broadcom, and Nvidia) who can
afford to design and produce SoCs.
Could this situation lead to a slowdown in semiconductor innovation?
Scott McGregor of Broadcom said,
"A new chip costs between US$30-
$40 million to develop. That is hard
for a VC to swallow. The number
of chips being developed is going
down, therefore the number of semiconductor companies should also
come down."
If the number of companies should
also come down, shouldn't the few
that remain shoulder the responsibility for funding the innovation? Can
they continue to rely on an acquisition
strategy for innovation?
Rappaport has expressed the view
that if the big semiconductor companies actually did the innovation
they ought to do, the VCs wouldn't
be needed to carry the load. He sees
little benefit to VC in subsidizing
the R&D for a large company. Rappaport showed the return multiples for
semiconductor companies that exited
between 2002 and 2011 (often fewer
than one). He said, "It's a dangerous
process. It is hard to get up in the
morning and say; Gee today I am going
to make an investment where I have a
50% chance of losing 60% of it." ... "We
are not getting paid for the risks we
take. So maybe we can figure out a way
to make these companies cheaper. But
the most interesting problems are very
complex to solve and this complexity
stretches product cycles making it
harder to take others by surprise".
T.J. Rodgers of Cypress Semiconductor later added, "I have some
of my money in VC. They have a

IEEE SOLID-STATE CIRCUITS MAGAZINE

responsibility to make money not
prop up an industry."
Rappaport concluded, "I admire
Dado for soldiering on as the last
man standing."
Banatao of Tallwood Venture Capital later humbly explained "It's really
all that I know." (Banatao manages
a VC fund that exclusively invests
in distinctive semiconductors and
semiconductor-related technologies
and products.)
Both Lu and McGregor would later
highlight alternate sources of funding
for semiconductor companies (other
than U.S.-based VC firms).

What Are VCs Looking for?
-Performance/Watt
If the VCs are moving away from
semiconductors, who will fund the
innovation to take the industry to the
next 20 years, and is such innovation
required? An audience member asked
if VC investment was moving away
from semiconductors to applications.
John Doerr answered, "Yes. Applications that run on smartphones
attract a lot of capital."
John added some perspective on
how far the industry has come.
"Seven years ago the iPhone didn't
exist. Four years ago, the iPad didn't
exist. These two products are now hundred of billions of revenue today. Let's
not lose sight of the rate of innovation
in this industry. Innovation really is
better now than ever before," he said.
In line with tremendous challenges and hence the continued need
for innovation that mobility and portable systems enjoyed by masses are
inducing on power efficiency (for
example in mobile computing, any

Table of Contents for the Digital Edition of IEEE Solid-State Circuits Magazine - Spring 2014