Kaku Book Review 3
Reviewing his book, Physics of the Future

Channel AskTheBots: http://www.youtube.com/watch?v=uIy1djjHWtk
Scene 1, College Campus
Jeff:
Hey, there's Susie over there! Hi, Susie. she probably cannot hear us from here.

Ian: Hi. My name is Ian.   I'm a new student here at Grove High. My family moved here from
England.

Jeff: Glad to meet you, Ian.  I like football...err...that would be American Football to you.
Let's walk down this way toward Susie.    Tell me more about what you are doing.

Ian: I like to read books, mostly non-fiction stuff. I just red Kaku's book "Physics of the Future".  
Boy, is that fellow strange.  He thinks that T-Shirts are stenciled, not silkscreened.  Of course,
some t-shirts are stencilled, but not in mass production. Then he went on to claim that in-teh-gra-
ted circuits are stencilled as well.

Jeff: Hasn't he ever visited a semiconductor facility?  

Ian: It appears that he has not.  He states that you take a stencil and put it over a silicon
wafer and etch the transistors with acid.

Jeff: Couldn't he just look it up on Wickepedia and get it right?

Ian: He must know about Wikipedia, maybe he doesn't use it.

Jeff: Here it is on my smartphone.  Processing steps are deposition, removal, patterning and
modification of electrical properties.  They use photo-masks, not stencils.

Ian: He thinks that Moore's Law is just about finished.  After 2020, Silicon Valley is likely to
become a rust belt, he says.

Jeff: Where does he live?

Ian: He teaches in New York, so I suppose that he lives there too.

Jeff: Okay, that makes sense. He is expressing his East coast rivalry hope that Silicon Valley
will become less dominant in the future.

Ian: Good luck to his hoping for the demise of Silicon Valley.  

Jeff: Oh, that is right, Michio Kaku, isn't he that pop-science guy that explains everything in
simple terms?

Ian: Yes, we are supposed to say "Golly gee, Mr. Science, tell us more".

Gail: Excuse me for chime ing in, guys.  My father heard Kaku speak at Microsoft and he
talked about the demise of Moore's law.  It was quite embarrassing and my father was worried
that people in the audience would start laughing.  Everyone was nice and polite, though and they
diverted their eyes as he embarrassed himself.

Ian: Cool information. I am Ian and this is Jeff.  What is your name?

Gail: I am Gail, pleased to meet you.  Got to run to class now.

Jeff: Pleased to meet you Gail.  Hey, [pause] let's walk where we can see Susie better.

Ian: Cool.

Jeff: So, you were saying that he thinks that Moore's Law is just about finished.  And after
2020, Silicon Valley is likely to become a rust belt?

Ian: And he actually says this before integrated circuits even start to utilize the third dimension.
 Intel is going with 3D transistors this year.

Jeff: My father is working on a design where 16 chips are scrubbed onto the same package.  
Just the chip contacts for the power and exterior edges are bonded out.

Ian: That sounds cool.  How do they get to the data on the interior chips?

Jeff: Oh, all the chips talk to each other with optical fibers which are integrated into the
chips.  Just take that one design with 16 chips in a package and it represents four doublings right
there and that translates to 8 more years of Moore's Law without any sweat.

Ian: But wouldn't they cost 16 times as much, given that there are 16 chips.

Jeff: Well, you could say that, but consider the price of today's chips in four years.  Their price
will be only one sixteenth of today's price.

Ian: That is a cool observation.  They will get 16 times the processing power and will be using
the old semi-fabrication facilities.  But won't that be a lot of power to dissapate?

Jeff: They are going with the lower power Atom chips.  Each chip is only 2 watts. They can
stack the chips as well. Eight years gets us to 2028 with Moore's Law.

Ian: How high can they stack them and still keep them cool?

Jeff: My dad says that they have been working on nanotube fins for years now.  They are
mechanically superior compared to other materials being ten times lighter, flexible, and stiff at the
same time. It appears that it will be easy to stack them four high and with some progress in
engineering, they will probably be stacking them much higher.

Scene 2
Ian:
So, you think that there are another 8 years at least just by stacking the chips.

Jeff: Yes.  That will get us to 2036.

Ian: I re[a]d that as the geometries get smaller, the R C capacitance between the wires becomes
more significant due to their increased proximity.

Jeff: So, they are going to have to do something to maintain the faster C P U clocks  as the
geometry gets smaller.

Ian: Yes, again, it looks like fibre optics will fill the bill.   Many small C P Us[e] and fibre optic inter
connects will get rid of the millimeter  long runs and allow very high clock rates.

Jeff: Sooner or later we have to move from silicon to carbon.   Back in 2010, they showed
graphene transistors operating at 100 giga hurts.  That carbon sure is fast.  And it is the best
conductor of heat as well.

Ian: So, you are saying that today's 4 giga hurts C P Use will be running at a  conservative 48
giga hurts or another factor of 8 times.

Jeff: Yes, and 8 represents 6 more years of Moore's Law. That gets us to 2044.

Ian: Just think.  With a diamond substrate it will be easy to use light to communicate instead of
electrical signals.  Wow, are things ever going to go faster.

Jeff: Yes, and scrubbing chips next to each other will be straight forward with the optical
communications made easy with diamonds.

Ian: And then there are the memristors that HP is going to ship starting in 2013.  My dad says
that they will be low power and similar to nerve cells.  With such low power, it seems very feasible
to stack them upwards and wider as well.  You might keep Moore's Law going to 2050 with such
breakthroughs.

Jeff: We have described how to get Moore's Law extended to 2050 and we have not even
considered new ideas which will be thought of in the future.

Ian: Looks like I have plenty of material for my book report, just on Kaku's Moore's Law
comments. Thanks a lot Jeff.

Jeff: Your certainly welcome. We didn't even touch on nanotechnology yet. Don't forget to run
me a copy of your book report.

Ian: Nice talking to you. I've got some pizza in my locker. Come join me.

Next
AskTheBot's Site for Discussion of Robotic Issues