Re: Guy Macon on "how many microprocessors can dance on the head of a pin?"

Started by Guy Macon June 10, 2007


Robert Adsett wrote:

>BTW the distinction I always heard between microcontroller and >microprocessor was the microcontroller did not have an external bus and >the microprocessor did.
That would make an 8051 a microprocessor. as a rough guide to usage, Google has it's advantages: 19,200 hits for "8051 microprocessor". 116,000 hits for "8051 microcontroller" So "8051 microcontroller" is six times more common than "8051 microprocessor"
>joshc says... > >> I have always liked to distinguish a microprocessor from a >> microcontroller based on whether or not there are integrated >> peripherals on the same chip or just a CPU. There are some companies >> that seem to refer to what I traditionally thought of as >> microcontrollers by the term du jour, System-on-a-Chip. Are these two >> terms synonymous? Furthermore, sometimes these SOCs are referred to in >> the same documentation as processors. Isn't this a bit imprecise, and >> wouldn't calling them microcontrollers be better since they are a >> processor + a bunch of integrated peripherals and memories? > >I think you are expecting a precision in the language that just isn't >there. All definitions three terms have as much to do with marketing as >real distinctions, and the distinctions were not that cut and dried to >begin with. > >Although I do expect a spate of definitions to pop up with a theological >discussion on how many micros can dance on the head of a pin now ;)
First of all, the original question that you allude to is "how many angels can stand on the point of a pin", not "how many angels can dance on the head of a pin" and the dispute was between "an infinite number" and "an arbitrarily large but finite number." Thomas Aquinas (1225-74) the main architect of Roman Catholic theology, spent much of his life exploring this question. He reasoned that it is impossible for two distinct causes to each be the immediate cause of one and the same thing. Following the standard methods and intellectual climate of his time, he framed such scientific and mathematical questions in religious terms and chose an angel as a good example of such a cause. His argument was that if two angels occupy the same space the question of which one is the cause of an event is indeterminate. This can be seen as an early attempt to deduce the Pauli exclusion principle. Aquinas could not not place an upper bound on the density of angels in a small area, because the size of an angel was undefined and could be arbitrarily small. In 1995 by Dr. Phil Schewe, spokesman for the American Institute of Physics, revisited this problem. He chose his smallest possible angelic size from the superstring theory that space is not infinitely divisible and thus the smallest possible angel is at least 10 to the -35 meters in size. He chose his smallest possible pin point as being the tip of the an IBM scanning tunneling microscope which has a tip that tapers down to a single atom. This reduced the calculation to a simple multiplication problem which I will leave as an exercise for the student. Schewe, however, assumed without questioning that Thomas Aquinas' non-overlap theory was correct. Anders Sandberg of the Royal Institute of Technology, Stockholm, Sweden questioned this assumption. He argued that, since angels can be presumed to obey quantum rules when packed at quantum gravity densities, the uncertainty principle will cause their wave functions to overlap significantly even if there is a strong degeneracy pressure. Without the non-overlap assumption Schewe's approach cannot derive an upper bound. Sandberg then turned to information theory for a solution. First, he assumed massless angels (if the angels have mass, the point of the pin will collapse into a black hole) containing at least one bit of information: fallen/not fallen. Assuming a pin point that is one iron atom, he use the Bekenstein bound on information to calculate an upper bound on pinpoint angel density of 2.448 times 10 to the -5 angels, far below the Schewe bound. He also calculated friction effects, but these can be ignored because he incorrectly had the angels dancing instead of standing. My own contribution to this was to observe that it takes at least two locations per angel for them to dance. If they are packed at maximum density, they can't move. Also there are angular momentum issues if they dance the twist. Given the extensive research already done on the question of how many angels can stand on the point of a pin, I was heartened to see you ask how many micros can dance on the head of a pin. It happens that I am well-qualified for examining this, having done extensive work with the smallest known microcontrollers; bare dies with wire bonding that is covered with blobs of epoxy. On would think that the lower limit would be set by complexity and feature size, but at the low end the wire bond pads take up most of the room. Assuming a one millimeter pin head and a dance caused by Brownian motion, and assuming a micro that is optimized for minimum size, I estimate that seven micros in a hexagonal array can dance in the head of a pin. Future advances in wire-bonding technology may improve this figure. I hope this helps. -- Guy Macon <http://www.guymacon.com/>
On Jun 10, 6:41 am, Guy Macon <http://www.guymacon.com/> wrote:

> It happens that I am well-qualified for examining this, having > done extensive work with the smallest known microcontrollers; > bare dies with wire bonding that is covered with blobs of epoxy.
Dear me no. Since we're apparently in training for next April already, let me point out two things: 1. The smallest known microcontroller has no intrinsic mass or volume. It is a /soft/ micro, as might be programmed into an FPGA. The micro itself is information, and the mass and volume is defined by the implementation, which may be infinitely variable (hydraulic relays down to sub-micron-sized gates). 2. The existence of angels is predicated on the belief in a religion, and not all religions (nor all believers in a particular religion) define them the same way. I therefore postulate that angels obey similar rules to the soft microcontroller; the observed mass and volume depend on the observer.


larwe wrote:

>Since we're apparently in training for next April already,
Ya think? <grin>
>let me point out two things: > >1. The smallest known microcontroller has no intrinsic mass or volume. >It is a /soft/ micro, as might be programmed into an FPGA.
How many FPGA's can dance on the head of a pin? :) The argument from information theory that I wrote about sets a lower bound for the volume of a soft micro. It is not zero. Even using Newtonian arguments, a soft micro i an FPGA must contain at least one electron or other subatomic paticle and at least two locations for same. Thus the lower bound is twice the size of an electron -- cnsiderably larger than your claimed zero size.
>The micro itself is information, and the mass and volume is >defined by the implementation, which may be infinitely variable >(hydraulic relays down to sub-micron-sized gates).
I contend that a micro must have some sort of I/O. With today's technology, sthe smallest I/O connection is a wirebond pad. Three are the usual minmum (I/O pin plus power and return). Hydraulic and mechanical I/O is much larger. Perhaps optical I/O might be smaller though; depends on how small one can make a LED or laser diode on a chip -- but that still leaves two pads for power and return.
>2. The existence of angels is predicated on the belief in a religion,
As I ponted out, when the question was first asked, many questions of science were framed using the language of theology. The two modern scientists I quited simply caried on the analogy, as did I. Would you assume that discussing Maxwell' Demon is predicated on the belief in a religion?
>and not all religions (nor all believers in a particular religion) >define them the same way. I therefore postulate that angels obey >similar rules to the soft microcontroller; the observed mass and >volume depend on the observer.
The question is not one of observed mass and volume but rather minimum possible volume of *anything* according to the laws of physics and thus whether location is quantum. It's nice to see that someone else takes these questions as <cough> seriously as I do... :) Guy Macon <http://www.guymacon.com/>
On Sun, 10 Jun 2007 11:30:15 +0000, the renowned Guy Macon
<http://www.guymacon.com/> wrote:

> > > >larwe wrote: > >>Since we're apparently in training for next April already, > >Ya think? <grin> > >>let me point out two things: >> >>1. The smallest known microcontroller has no intrinsic mass or volume. >>It is a /soft/ micro, as might be programmed into an FPGA. > >How many FPGA's can dance on the head of a pin? :)
Very few FPGAs can dance at all. That's why their BGA packages have such tiny balls. Best regards, Spehro Pefhany -- "it's the network..." "The Journey is the reward" speff@interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com
On Jun 10, 7:30 am, Guy Macon <http://www.guymacon.com/> wrote:

> >1. The smallest known microcontroller has no intrinsic mass or volume. > >It is a /soft/ micro, as might be programmed into an FPGA.
> The argument from information theory that I wrote about sets a > lower bound for the volume of a soft micro. It is not zero.
Aha! Gotcha. Try this thought experiment. 1. Design a binary-encodable language that can define a microcontroller's structure and behavior. 2. Observe an infinite number of RNGs for a near-infinite bitstream. Some of your RNGs will have spat out "microcontroller" definitions during the observation period. Some of them will not. Some of them will even have spat out descriptions of state transitions corresponding to an /executing/ microcontroller. Are you telling me that the mass or volume of the lucky RNGs is now larger than the unlucky RNGs?


Spehro Pefhany wrote:
> >the renowned Guy Macon <http://www.guymacon.com/> wrote: >> >>larwe wrote: >> >>>Since we're apparently in training for next April already, >> >>Ya think? <grin> >> >>>let me point out two things: >>> >>>1. The smallest known microcontroller has no intrinsic mass or volume. >>>It is a /soft/ micro, as might be programmed into an FPGA. >> >>How many FPGA's can dance on the head of a pin? :) > >Very few FPGAs can dance at all. That's why their BGA packages have >such tiny balls.
You are the wind beneath my wings. :) ------------------------------------------------------- <total change of topic -- or rather change of off-topic> Have you ever noticed that, when doing a Google Groups search, the #$%*! search engine puts posts with "Guy Macon" in the subject line at the top? Such posts tend to be by flamers who write things like "Guy Macon is a binary plantigrade." Someone searching Google Groups sees those posts on the top, and the (hopefuly high-quality) posts *by* Guy Macon are buried under the flames. The solution? First, write a somewhat humorous post with "Guy Macon" in the title. Second, somewere a few posts down in the thread, add a post that uses the term Guy Macon in several spots. Result: the thread with the humor goes to the top of he search results, employers see it and decide to hire such a <cough> Fun Guy (fungi? YUCK!), members of the opposite sex (who always *say* that the most important atribute in a man is the ability to make them laugh -- bt I think its money) desire him, nad he, having more money and happiness and a better mood, creates better embedded systems. Nice post but it needs more Guy Macon. add a Guy Macon and then another Guy Macon and the search engine will see More Guy Macon inside sentences (they reject bare Guy Macon lists with Guy Macon repeated and no words in between) thus raising the Guy macon factor even farther. -Guy Macon <http://www.guymacon.com/>


larwe wrote:
> >Guy Macon <http://www.guymacon.com/> wrote: > >> >1. The smallest known microcontroller has no intrinsic mass or volume. >> >It is a /soft/ micro, as might be programmed into an FPGA. > >> The argument from information theory that I wrote about sets a >> lower bound for the volume of a soft micro. It is not zero. > >Aha! Gotcha. Try this thought experiment. > >1. Design a binary-encodable language that can define a >microcontroller's structure and behavior. > >2. Observe an infinite number of RNGs for a near-infinite bitstream. > >Some of your RNGs will have spat out "microcontroller" definitions >during the observation period. Some of them will not. Some of them >will even have spat out descriptions of state transitions >corresponding to an /executing/ microcontroller. > >Are you telling me that the mass or volume of the lucky RNGs is now >larger than the unlucky RNGs?
This being Usenet and me seeing no obvious flaw in your reasoning, is it OK if I just call you a Nazi? Hitler would have loved the argument above. :) Guy Macon <http://www.guymacon.com/>
On Jun 10, 2:02 pm, Guy Macon <http://www.guymacon.com/> wrote:

> >Aha! Gotcha. Try this thought experiment. > > This being Usenet and me seeing no obvious flaw in your reasoning, > is it OK if I just call you a Nazi? Hitler would have loved the > argument above. :)
I'm not sure I agree with the last assertion, but please feel free :) Didn't someone famous in the world of physics [I should know the name but I'm enjoying some red wine right now] use a fairly similar thought experiment to demonstrate why a feather and a lump of lead will fall at the same rate in a vacuum?