Physics Lournal

The Physical Nature of Information

Metadata

Author:: Rolf Landauer

Abstract: Information is fundamental to physical representation, which means it must also be fundamental to restrictions, and the (im)possible, with regards to the laws of physics.

Quantum Mechanics entertains the notion of Superposition, with regards to states (of systems) that bear information, and the real utility of this must be understood. Parallelism in Quantum Computation is one potential application, and will be critiqued.

Computation requires the dissipation of energy, as does measurement, and the communication link.

1. Information is physical

It would be incorrect to think of information, as an abstract and disembodied substance: there will always be some physical representational aspect to information. Information must be represented by something: a physical substrate is the implication, but more greatly, it can not be represented by something that is above the laws of nature.

We see reference to this in Leo Szilard's contemplation of Maxwell's Daemon, however this did not end the debate about the nature of information, and its physicality. Roger Penrose, in a rather Platonist, manner, upholds that "...devices can yield only approximations to a structure that has a deep and 'computer-independent' existence of its own.

2. Quantum Information

The first type of counting we became familiar with as a species was done on hands, with shells, and rocks. It was binary, leading us to indicate the existence of something as 'existent (1)', or 'non-existent (0). QM, however, allows for something to possess, or rather, be, in both states simultaneously.

One situation that draws on this, regards Quantum Teleportation. Given two entangled Einstein-Paul-Rosen objects, where one is shipped to a transmitting end, the other two a receiving end, interaction between the object at the transmitting end, and the source object, whose state is being transmitted, generates a classical signal. This signal, interacts with the prepared object, resulting in a copy of the teleported, or transmitted state.

The most likely application of this phenomena lies within the realm of Quantum Cryptography, which is currently achievable in real systems, although not efficient enough for it to be applicable in a widespread manner. It directly relies on The Indeterminacy Relation.

A stream or quantity of quantum information can not be examined by an eavesdropper, without altering it: examination is measurement, and eavesdroppers are observers.

Communication of bits requires less handling than computational processes that use them.

Additionally, errors in communication channels are easily handled with redundancy techniques.

If I remember correctly, there's an open problem in Quantum Information Theory regarding Quantum Repeaters and their current perceived infeasibility with regards to implementation.

It was Paul Benioff who first realized that electron Spin under quantum mechanical time evolution (h/t Schroedinger Equation) could be used for calculation just as classical bits were.

David Deutsch followed up on this, realizing that such a computer could follow, as opposed to a program, a superposition of different computational pathways in parallel, and at the end of the "programs" lifetime, we would have information that depended on all of these pathways.

Peter Shor then proved that this type of parallelism could in fact provide performance gains for factoring large numbers, which was a huge milestone.

Quantum Parallelism: A Return To Analog Computation

An analog computer is more capable than a digital one, however the nature of physical phenomena, such as voltage, that can take on any value from a large set of potential values, is not exactly amenable to Error Correction.

Due to the nature of this kind of computing, imperfections build up quickly, leading to programs that can only complete a few successive tasks before these errors begin to effect the quality of the computation and results.

Digital computers however, only allow binary values, which permits easier restoration of signals before they become too corrupted to be saved. This is what allows digital computers to carry out massive numbers of computations successively, with little to no notable errors.

Within quantum parallelism, we use all possible superpositions of 1 and 0, which is the source of the power of the technique, but it also injects the problems of analog computing back into the situation.

Energy Dissipation Requirements

Impact On The Laws of Physics.

Referenced in

Rolf Landauer

The Physical Nature of Information

Does Time Really Flow?

The problem arising here is that information has a physical aspect, and requires energy and space for its existence.

The Physical Nature of Information

The Physical Nature of Information

Metadata

Abstract: Information is fundamental to physical representation, which means it must also be fundamental to restrictions, and the (im)possible, with regards to the laws of physics.

Quantum Mechanics entertains the notion of Superposition, with regards to states (of systems) that bear information, and the real utility of this must be understood. Parallelism in Quantum Computation is one potential application, and will be critiqued.

Computation requires the dissipation of energy, as does measurement, and the communication link.

1. Information is physical

2. Quantum Information

The most likely application of this phenomena lies within the realm of Quantum Cryptography, which is currently achievable in real systems, although not efficient enough for it to be applicable in a widespread manner. It directly relies on The Indeterminacy Relation.

A stream or quantity of quantum information can not be examined by an eavesdropper, without altering it: examination is measurement, and eavesdroppers are observers.

Communication of bits requires less handling than computational processes that use them.

Additionally, errors in communication channels are easily handled with redundancy techniques.

If I remember correctly, there's an open problem in Quantum Information Theory regarding Quantum Repeaters and their current perceived infeasibility with regards to implementation.

It was Paul Benioff who first realized that electron Spin under quantum mechanical time evolution (h/t Schroedinger Equation) could be used for calculation just as classical bits were.

David Deutsch followed up on this, realizing that such a computer could follow, as opposed to a program, a superposition of different computational pathways in parallel, and at the end of the "programs" lifetime, we would have information that depended on *all* of these pathways.

Peter Shor then proved that this type of parallelism could in fact provide performance gains for factoring large numbers, which was a huge milestone.

Quantum Parallelism: A Return To Analog Computation

An analog computer is more capable than a digital one, however the nature of physical phenomena, such as voltage, that can take on any value from a large set of potential values, is not exactly amenable to Error Correction.

Due to the nature of this kind of computing, imperfections build up quickly, leading to programs that can only complete a few successive tasks before these errors begin to effect the quality of the computation and results.

Digital computers however, only allow binary values, which permits easier restoration of signals before they become too corrupted to be saved. This is what allows digital computers to carry out massive numbers of computations successively, with little to no notable errors.

Within quantum parallelism, we use all possible superpositions of 1 and 0, which is the source of the power of the technique, but it also injects the problems of analog computing back into the situation.

Energy Dissipation Requirements

Impact On The Laws of Physics.

Referenced in

Rolf Landauer

The Physical Nature of Information

Does Time Really Flow?

David Deutsch followed up on this, realizing that such a computer could follow, as opposed to a program, a superposition of different computational pathways in parallel, and at the end of the "programs" lifetime, we would have information that depended on all of these pathways.