Red Herring on Andrew Lines' new company, ADD

Date view Thread view Subject view Author view

From: Rohit Khare (
Date: Mon Jun 19 2000 - 20:34:32 PDT

Lab Rat: Going clockless
By Phil Harvey, June 15, 2000

                         It's tough for university-bred technology, no
matter how novel, to
                         grab the attention of the business community,
let alone the
                         general public. But take that same sleepy
technology and wrap a
                         youthful startup around it, and it won't be
long until the smell of
                         money awakens investors, the press, and the
man on the street.

                         That's what's happening to a tiny Pasadena,
                         startup called Asynchronous Digital Design
(ADD). The firm says
                         its technology -- based on a decade of
research from the
                         California Institute of Technology -- will
help it create a new kind
                         of high-performance microprocessor that
doesn't need a clock,
                         one of the required ingredients in most of
today's chips.

                         According to the company's founders, the
radical design could
                         produce semiconductors that have the same
performance level of
                         today's top chips, while consuming less
energy and having a
                         longer shelf life. Although the hurdle of
bringing such a product to
                         market is unprecedented, the growing need for
more low-power
                         microprocessors and a possible end of the
road for Moore's Law
                         makes ADD an idea worth considering.

                         Barely out of the dorm, ADD's attracted some
early funding from
                         the likes of Yobie Benjamin, Ernst & Young's
top strategist for its
                         e-commerce and emerging technology practice.
The firm has also
                         received investments from other individuals
who aren't quite as
                         chatty as Mr. Benjamin is during a recent
happy-hour meeting at
                         San Francisco's Mars Bar.

                         MOORE'S LAW VS. PHYSICS
                         In any case, ADD's development will be
interesting to watch in the
                         next few months because its technology and
timing seem to be
                         spot-on with the growing trend of computing
devices becoming
                         smaller and mobile, putting a premium on
speed and power

                         Not many specific details about ADD's chip
are available yet. "We
                         don't want to talk about [the] processor
until we have some
                         silicon fabbed," says ADD President Uri Cummings. Mr.
                         Cummings, who founded the firm with Caltech
classmate Andrew
                         Lines, was nice enough, though, to help
explain why his firm's
                         asynchronous chip might someday be a big deal.

                         Since 1965, Moore's Law has always been an
easy measurement
                         of progress in the semiconductor industry. It
hails from Intel
                         (Nasdaq: INTC) co-founder Gordon Moore's
classic observation
                         that every 18 to 24 months, the number of
transistors on a chip,
                         and the chip's performance, will double.

                         For the most part, semiconductors have
evolved right along that
                         path. Today's chips are faster, and they use
more power than
                         their predecessors do. But the faster the
chips are, the tougher
                         and more expensive they are to design, build,
and maintain. And
                         because transistors give off more heat as
more are crowded onto
                         a chip, it's reasonable to assume that they
won't be able to
                         progress beyond certain physical limits.

                         CLOCKLESS CHIPS
                         So what happens in, say, the next 15 years,
when chips continue
                         shrinking and giving off heat -- do they
finally get as small and
                         as fast as they can get? Mr. Cummings and
other proponents of
                         asynchronous chip technology have an answer
on the tips of their
                         tongues: Try chips without clocks.

                         Most processors today are synchronous, meaning a clock
                         regulates their internal timing. Today we
measure how fast a
                         computer can execute instructions by the
number of clock cycles
                         per second. (A 300-MHz processor, like the
one in my PC, carries
                         out 300 million cycles per second.)

                         Unfortunately, clock-driven chips waste a lot
of the power they
                         consume by running in place. In muddier
language, each part of
                         the clock-governed processor operates in lock
step, meaning that
                         all stages in the processor's operation take
exactly as long as the
                         slowest stage. Instructions are passed from
each chip subsystem
                         to the next like cars on an assembly line.

                         ROCK AROUND THE CLOCK
                         Contrast that with asynchronous chips, where
the transistors can
                         theoretically switch independently of one
another and run at their
                         own speed. An asynchronous chip's subsystems can swap
                         information at mutually negotiated times,
without having to check
                         in with a clock. As a result, an asynchronous
chip exhibits the
                         power of its components' average performance.
A synchronous
                         chip, however, exhibits the performance of
its slowest component.

                         Also, an asynchronous chip also saves power
because it can shut
                         down parts of the circuit that aren't in use.
In other words, there's
                         no running in place.

                         Still, chip designers understand synchronous
chips better. Its
                         lock-step process seems a bit strange to
laymen, but its much
                         easier to visualize. PC Week's Peter Coffee
gave one of the best
                         illustrations of synchronous logic when he
wrote, "Imagine
                         working in an office where everyone picks up
the phone every 60
                         seconds to see if anyone is calling."

                         In turn, Byte Magazine's Dick Pountain,
described asynchronous
                         logic as a taxi service where cabs depart,
not at fixed times, but
                         only when they're carrying passengers.

                         Whatever you compare their craft with, Mr.
Lines and Mr.
                         Cummings are an interesting duo.
Theoretically, asynchronous
                         chips are the toughest kind to design. But
after years of studying
                         under Dr. Alain Martin, who developed and
"fabbed" the first
                         delay-insensitive, RISC-like asynchronous
processor in 1989,
                         these gentlemen say that asynchronous design
is the only kind
                         of circuit methodology they've been immersed in.

                         IN GOOD COMPANY
                         Naturally, they're not the only ones on Earth
studying this kind of
                         processor design. A good list of other firms
and universities
                         involved in this arena can be found at the
University of
                         Manchester's Web site.

                         The asynchronous backers seemed to be a
vibrant, somewhat
                         incestuous, technology community. As an
example, Dr. Ivan
                         Sutherland, a VP and Fellow at Sun Labs
(Nasdaq: SUNW), is well
                         known for his continuing work on asynchronous
processors. Mr.
                         Sutherland, who had lunch with the lads at
ADD earlier this week,
                         helped establish their field of study at
Caltech. It was during Dr.
                         Sutherland's teaching tenure at Caltech from
1976 to 1980 that
                         integrated circuit design became an
acceptable field of academic

                         Mr. Lines and Mr. Cummings hope their lack of
legacy in the
                         synchronous chip world will help them create
an asynchronous
                         chip that's low-power, low-cost, and capable
of running complex
                         software programs. Although the commercial
uses for their
                         processor, called Vortex, haven't been fleshed out, Mr.
                         Cummings says the first chip fabbed by ADD will perform
                         comparably to a traditional 1-GHz processor.
At that speed,
                         potential uses for the low-power chip could
be as diverse as
                         mobile communications products, portable computers, and
                         networking gear.

                         Although ADD seems to be off to a flying
start, it has a long way
                         to go before it reaches commercial success.
As things stand now,
                         many of the problems previous asynchronous
designs addressed
                         are solved well enough with today's
synchronous chips. For most
                         applications, manufacturers have found ways to make
                         synchronous chips that are fast enough
without hogging too much

                         Also, most semiconductor companies employ
those that have
                         been working in synchronous designs for
years. You can expect
                         that they'd want to protect their vested
interests, at least until
                         they've found a way to compete with or
subsume the younger

                         But as the world leans toward computing with portable,
                         Net-connected devices, asynchronous designs
could win big, or at
                         least become an interesting alternative in
the future. Of course,
                         this speculation depends on what will happen
when ADD makes
                         Vortex commercially available sometime next year.

                         Discuss today's Lab Rat column in the Labrat
column discussion.
                         Or visit the Discussions home page.

Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Mon Jun 19 2000 - 19:38:28 PDT