The Physics of Information Technology

Danny Yee wrote this review for everyone who likes reading physics textbooks. Danny says: "If you studied physics at uni and are interested in a different kind of text at that level, read on. If not, you might prefer my review of Measured Tones: the Interplay of Physics and Music or other popular science reviews. The Physics of Information Technology author Neil Gershenfeld pages 370 publisher Cambridge University Press rating 9 reviewer Danny Yee ISBN 0-521-58044-7 summary Dense but rewarding

The Physics of Information Technology is a physics text, not a work of popular science: it assumes the reader has done a physics degree or the larger part of one. The connection with information technology is threefold: Gershenfeld takes an information-theoretic approach at a fundamental level, focuses on areas of physics relevant to information technology, and uses examples from computing systems. The result is dense but richly rewarding, covering an immense range of material and often providing a different perspective on it to that of more traditional physics textbooks. (The Physics of Information Technology might be suitable as a text for an advanced electrical engineering course.) Enhancing the work's utility for students, each chapter has a "selected references" section, which lists maybe half a dozen books along with one sentence descriptions, and a set of problems, with full worked solutions.

Gershenfeld starts with chapters on noise and information in physical systems, covering noise mechanisms, the equipartition and fluctuation-dissipation theorems, channels, Shannon's theorems, and Fisher information. A rapid electromagnetism refresher is followed by a chapter on circuits, transmission lines, and waveguides, and another on antennas. A general review of optics is followed by a chapter "Lensless Imaging and Inverse Problems", covering matched filters, coherent imaging, computed tomography, and magnetic resonance imaging. Turning towards solid state physics, a quick overview of quantum statistical mechanics and electronic structure leads to an explanation of the operation of junctions, diodes, and transistors and various kinds of semiconductor logics; a chapter on opto-electronics looks at systems for the generation, detection, and modulation of light; and a chapter covers magnetic materials and recording. Two chapters then link this back to the information theory, covering instrumentation and signal modulation, detection, and coding and, adding complexity, many-body effects (superconductivity), non-equilibrium thermodynamics (thermo- and piezo-electricity), and relativity. And a long final chapter offers a solid introduction to quantum computing and communications, starting with an explanation of the necessary quantum mechanics.

Gershenfeld packs a huge amount into The Physics of Information Technology. Though he does review background theory, he does so rapidly and then cuts straight to the essentials. The section on coding, for example, explains arithmetic and Huffmann compression in just a paragraph each, while two and a half pages on thermoelectricity explain thermocouples and Peltier coolers. The mathematics is perhaps an exception, with the bits Gershenfeld chooses to treat in detail (and it gets quite involved in places) sometimes rather arbitrary - the mathematics can usually be skipped without too much loss. So the discussion of ferro- and ferri-magnetism includes a page and a half of mathematics deriving the Heisenberg Hamiltonian and J coupling, but then drops out of "mathematics mode" pretty much entirely (with one paragraph here quoted as an example of the style):

"In an antiferromagnet such as Mn or Cr the exchange energy is negative, therefore neighbouring spins alternate orientation and there is no net movement even though there is long-range magnetic order. A ferrimagnet is a ceramic oxide that has a spontaneous moment but is a good insulator. The moment arises because it has an antiferromagnetic coupling, but there are interpenetrating spin-up and spin-down lattices that have different moments but do not cancel. Most common ferrimagnets are made from materials containing iron oxides, called ferrites. Because they do not conduct, they do not screen electric fields or have eddy current heading, and so they are useful for a range of microwave applications as well as guiding flux in coils. One example is the microwave equivalent of optical Faraday rotation, which is used in a "magic T" to steer microwave signals in different directions depending on whether they arrive at the input or the output port. This apparent violation of reversability is possible because magnetic interactions break time reversal invariance, since the sign of time appears in the velocity in the basic vxB law. Cables are often wrapped around ferrites, such as the beads on computer monitor cables, to add inductance to filter out unwanted high-frequency components."

This also illustrates the use of examples from computer hardware.

Table of Contents:

1. Introduction
2. Interactions, units, and magnitudes
3. Noise in physical systems
4. Information in physical systems
5. Electromagnetic fields and waves
6. Circuits, transmission lines, and wave guides
7. Antennas
8. Optics
9. Lensless imaging and inverse problems
10. Semiconductor materials and devices
11. Generating, detecting, and modulating light
12. Magnetic storage
13. Measurement and coding
14. Transducers
15. Quantum computing and communications

Purchase this book from FatBrain. Danny Yee has written nearly six hundred book reviews.

Comments

Was just joking

[ch-chuck]

with someone the other day about how our products obey the laws of physics, but software, GEEZUS! Software obey's no laws whatsoever! Neither natural nor social.

Re:Was just joking

[MindStalker]

Well it does obey the laws of logic. Of course you can program it to seem illogical, but ultimatly it is still logic, just very strange logic.

Re:Was just joking

[snatchitup]

This book sounds like an overview of 1/4 of my E.E. degree.

I think the most relevant, and or, widely misunderstood things is the theory of what information actually is in terms if bits. Which basically gives a theoretical limit on how much we can compress things.

Re:Was just joking

[eh2o]

i.e. shannons law, nyquist, etc.-- widely misunderstood perhaps by the layman -- but really I think this is a very simple concept.

btw-
Gershenfeld's _Nature of Mathematical Modeling_ is another book to check out... its more numerical analysis/computational modeling oriented, and like _Physics of Information..._ its also incredibly dense.

I would not recommend Gershenfeld's books unless one already posesses a good deal of knowledge about the topic... the presentation is too abridged.

Re:Was just joking

[Anonymous Coward]

Several intel chips have exhibited non-deterministic behaviour due to hardware bugs- not on purpose of course, but could introduce true illogicality into your programs, like, say if you're doing a conditional test on a certain floating point value in an early Pentium :-)

Also

[milo_Gwalthny]

If you're interested in this topic, you may also want to check out "Feynman Lectures on Computation" (Paperback - 320 pages, July 2000
Perseus Books; ISBN: 0738202967). Less hardcore, and the semiconductor physics stuff is dated, but everyhing explained in that great Feynman way.

Here's my contribution to the physics of IT:

[Rudeboy777]

Sitting in a chair for > 8 hours a day leads to an increase in mass and a decrease in kinetic energy.

Re:Here's my contribution to the physics of IT:

[Anonymous Coward]

Sitting in a chair for > 8 hours a day leads to an increase in mass and a decrease in kinetic energy.

But think of all the potential energy stored in the fat!

Re:Here's my contribution to the physics of IT:

[Dwonis]

Depends. Do you mean in the fat molecules, or the atoms?

Related research - Splitting the Octave

[Anonymous Coward]

I ran across this in a google search:

http://www.dgp.toronto.edu/~ematias/papers/music

It's pretty easy to grasp and (unlike the book) it's free! :-)

More Storage

[grepnyc]

So there's more to it than just trying to find space for more storage devices?

I work for a financial company who is a big IBM mainframe customer.... the guys who maintain the frames' are always looking for places to keep the DASD packs.

This problem even scales down to us distributed systems guys... where do we put the raid?? There's no more room for new servers!!! There's no more room in that tower for another drive!!!

Perhaps this book can help?

Physics?

[clandaith]

I thought the physics of networking only had 3 parts:

1. Computer connected to a networking line.

2. Network line connected to the WAN cloud.

3. Another network line coming out of the WAN cloud to another computer.

You meant there is more to it than just that!!

Re:The Physics of Information Technology

[Anonymous Coward]

Umm, yeah . . . and the rest of slashdot is different how ?

Is this really relevant?

[Limburgher]

I mean, this is all wonderful, but when will they increase the maximum network length for Token Ring lobes so I can run a direct line across town to my friend's house for 16Mbit interactive gaming heaven? ;)P

And more: physics of wireless information transfer

[mjackson14609]

How scattering can create additional usable channels, from the current Physics Today:

http://physicstoday.org/pt/vol-54/iss-9/p38.html

Sounds Good

[morbid]

This sounds like a really good read. It's all very well being able to dump bytes to a device or read packets from the network, but a good understanding of what's going on in a physical sense is invaluable.
Maybe one of these days when I've read the other 2 dozen books I have lying around waiting, I'll get this one.
With this sort of understanding, great things should be possible.
Please post more reviews of proper text books such as these.
PS Those of you in the UK, in case you didn't already know, your local public library is obliged to get you any book you ask for, no matter how obscure. It's the law. So get this one :-)

Re:Sounds Good

[GCP]

Well, I'm an American, but I like you Brits, so might I suggest an alternative? Buy the book, read it, then donate it to your local UK public library, and let them use their limited budgets for other things?

Just my two cent^H^H^H^H^H^tuppence.

Book information?

[Liam]

It would be nice if the author had given basic bibliographic information up front - title, author, publisher. It's customary in book reviews.

Good reading about the physics of tones

[more]

The article mentioned one book about physics of tones. I have one definite favorite in this area. Actually it is more psychoacustics (physics of the basilar membrane in the inner ear) than just good old traditional physics...

If you really want to understand tuning and how it is connected with spectra of sounds you should read Bill Sethares excellent book "Tuning, Timbre, Spectrum, Scale". Take a look at this article to get a preview about what the book is all about. He is not using the concept of a harmonic template at all, but relies solely on sensory dissonance (by Plompt and Levelt). The results are still quite usable in composing music.

IMHO, this book is about the only way the usual geek can understand the basics of harmony, consonance, and composition.

Re:two days ago my girl went "SLURRRP" with my cum

[Anonymous Coward]

your girlfriend is a cumslurping whore, HA HA