2 - 3 -
4 - 5
- 6 -
7 - 8
- 9 -
10 - 11 -
12 - 13
Speech at the Society for Neuroscience, November 4, 2007
Public Advocacy Forum: Translating Neuroscience: Can
Systems Engineering and
Lessons from High-Tech Take Us Beyond the
San Diego Convention Center: Room 4
Date: Sunday, November 4, 2007
John H. Morrison, PhD
"Andy Grove, former CEO of Intel and Time magazine's Man of the Year in 1997, is a strong advocate for accelerating the pace of biomedical research. Mr. Grove discussed the limitations of the current approach to bioscience research, and argues that to effectively translate basic neuroscience into new ways to fight brain disorders requires a “cultural revolution” in the research community and a major rebalancing of NIH spending. ."
Intro by John Morrison,
Society for Neuroscience
Andy Grove: One
correction and one comment, please.
John Morrison: Oh no.
Andy Grove: I went to
City College during its first golden era.
John: Oh okay, when it
was the goldest of the gold.
Andy Grove: The second one is a request by one of my associates reflecting on the power of technology. For those of you that happen to know the score of the Colts game, keep it to yourselves because he’s time shifting and wants to see it new.
The two industries that I'll talk about, I had the opportunity to participate in, one voluntarily, one less so. The one that I participated in voluntarily is the semiconductor industry that fuels today’s electronics, and it's a scientific industry and it was probably as rudimentary at the time I joined it as some of the fields that many of you are working on today. And it has grown up to be a $250 billion dollar industry fueling other industries that are probably about a trillion dollars in size. The other industry, bio-enterprise, I don't quite know what to call it, it's not an industry, it's academia, government and commercial entities making up something twice the size, including biotech and pharmaceuticals. I don't quite know how to define the boundaries, but it's a large, large industry. It's also a science-based industry. I got to become a student and participant 12-13 years ago when I encountered my first induction, enlistment or draft. I had prostate cancer and my involvement started with that. I became familiar with cancer research. Not to leave good enough alone, came Parkinson’s some years later. Those are my credentials for participation.
What I will base my talk on are some of the comparative factors, similarities and differences. Both of them are science-based industries, both of them are large and without attempting to blow smoke, both of them are peopled with highly competent, educated and dedicated individuals. I can walk the halls of this convention center and feel at home, with the one possible exception there are more women in this group than in technology. That's a strike against technology.
With that the differences
start. Let me start by giving you a thumbnail sketch of my technology
career, which started when the wafers were two-inch size and I had hair.
Forty years had a
considerable impact on me.
(photo of ASG 1968 morphs
into present-day ASG)
I'll leave it as an
exercise for the audience to decide whether it's a positive or a
But it had a comparable
impact on the technology as signified by the change in the size of the
(ASG holds up actual 12”
That's today's production
wafer. Which changed more?
That is one element of productivity, larger silicon wafers, and by the way, the complexities of making single-crystal silicon chunks, (they are called ingots), from which this wafer is cut are incredible. Other elements had to do with putting more and more components on every little chip that comes out of a silicon wafer, and this has been a driving force of the industry since the 1960's. It's called Moore's Law, named after Gordon Moore, a colleague and boss of mine, from even before Intel started. And, the empirical observation is that the number of components we can put on a silicon chip doubles in approximately 18 months to two years. It's a sort of self-predicting, driving force because since the semilog plots never lie, we had to work to make it come true.
helpful to keep this going, as shown in the slide, describing the
growth of these components on a semilog plot, with microprocessors, from
the first one introduced in 1971 to the current one which has multiple
billions of transistors on each chip. That is what the industry has done
in terms of component density, and the consequences of it is that the
price per performance unit, dollars per millions of instructions per
second, is asymptotically approaching zero.
2 - 3 -
4 - 5
- 6 -
7 - 8
- 9 -
10 - 11 -
12 - 13