Simplicity and quantum complexity

May 04, 2012
Simplicity and quantum complexity
A simulator for a stochastic process can be thought of a physical system that stores select information about past outputs, and uses them to generate the require statistics for the future

Simulations of reality would require less memory on a quantum computer than on a classical computer, new research from scientists at the University of Bristol, published in Nature Communications, has shown.

The study by Dr Karoline Wiesner from the School of Mathematics and Centre for Complexity Sciences, together with researchers from the Centre for in Singapore, demonstrates a new way in which computers based on quantum physics could beat the performance of classical computers.

When confronted with a complicated system, scientists typically strive to identify underlying simplicity which is then articulated as natural laws and fundamental principles. However, complex systems often seem immune to this approach, making it difficult to extract underlying principles.

Researchers have discovered that complex systems can be less complex than originally thought if they allow to help: quantum models of complex systems are simpler and predict their behaviour more efficiently than classical models.

A good measure of the complexity of a particular system or process is how predictable it is. For example, the outcome of a fair coin toss is inherently unpredictable and any resources (beyond a random guess) spent on predicting it would be wasted. Therefore, the complexity of such a process is zero.

Other systems are quite different, for example neural spike sequences (which indicate how sensory and other information is represented in the brain) or protein conformational dynamics (how proteins – the molecules that facilitate biological functions – undergo structural rearrangement). These systems have memory and are predictable to some extent; they are more complex than a coin toss.

The operation of such in many organisms is based on a simulation of reality. This simulation allows the organism to predict and thus react to the environment around it. However, if quantum dynamics can be exploited to make identical predictions with less memory, then such systems need not be as complex as originally thought.

Dr Wiesner added: “On a more fundamental level, we found that the efficiency of prediction still does not reach the lower bound given by the principles of thermodynamics – there is room for improvement. This might hint at a source of temporal asymmetry within the framework of quantum mechanics; that it is fundamentally impossible to simulate certain observable statistics reversibly and hence with perfect efficiency.”

Explore further: 'Cavity protection effect' helps to conserve quantum information

More information: Nature Communications 3, 762 (2012). Abstract. arXiv:1102.1994.

Related Stories

How quantum physics could make 'The Matrix' more efficient

Mar 29, 2012

Researchers have discovered a new way in which computers based on quantum physics could beat the performance of classical computers. The work, by researchers based in Singapore and the UK, implies that a Matrix-like ...

Recommended for you

User comments : 11

Adjust slider to filter visible comments by rank

Display comments: newest first

Origin
1 / 5 (1) May 04, 2012
The von Neumann architecture of classical computers is sequential, the quantum computing is based on paralellistic approach - so it's logical it's less exact but it requires lower amount of memory. But can the computational power of quantum computers ever beat the classical computers at the moment, when the classical computers are already hitting the barrier of Heissenberg's uncertainty principle? For me it seems it cannot - the quantum computers could be very fast, but very approximate too at the price.
kochevnik
not rated yet May 04, 2012
Classical computers are still very, very far from operating at the Plank scale. And they can't work beyond it.
simplicio
5 / 5 (1) May 04, 2012
the quantum computers could be very fast, but very approximate too at the price

What do you mean "approximate"? You are confusing quantum computer with inaccurate commuter. Quantum computer can produce exact result or probability result depending on problem, just like classical computer.
Terriva
1 / 5 (2) May 05, 2012
Quantum computer can produce exact result or probability result depending on problem, just like classical computer.
Maybe they can, but the results of contemporary quantum computers are approximate only due the quantum noise. For example the Groower's algorithm has been reproduced with 96% reliability. But von Neuman computer cannot work with such level of noise at all.
Terriva
1 / 5 (2) May 05, 2012
For example, for quantum computer working with reliability 96% like the 64-bit classical computer you will need to repeat the computation 10^9 times to achieve the result of the same level of precision. Such a precision could be achieved with parallelizing of results of 10^9 atoms in the role of qubits. We can realize fast, such level of parallelization of quantum bits is just achieved in classical transistors, which are constructed from 10^9 atoms each. For me the quantum computers are just hype, the main reason of it is to provide the employment for researchers involved. But can they really beat the classical computers in computational power? IMO they cannot, but this power is limited with the very same laws both for classical, both for quantum computers.
antialias_physorg
5 / 5 (1) May 06, 2012
Comparing quantum computers to classical computers is comparing apples to oranges. They are both suited to solve entirely different sets of problems.

Quantum computers could give you the prime number factorization of large numbers alsmots instantly - while regular computers would slave over this for many lifetimes of the universe. On the other hand regular computers are well suited for deterministic calculations - which quantum computers have a hard time with.

The two complement each other - and in the end we'll probably see computers with a mix of regular and quantum cores.
Terriva
1 / 5 (2) May 06, 2012
Comparing quantum computers to classical computers is comparing apples to oranges

Seed of apples, being more specific - as we have no practical applications of quantum computers developed yet.
kaasinees
1 / 5 (1) May 06, 2012
Comparing quantum computers to classical computers is comparing apples to oranges

Seed of apples, being more specific - as we have no practical applications of quantum computers developed yet.

Artificial Intelligence
Some 3D calculations maybe.

Just thinking out loud.
El_Nose
not rated yet May 07, 2012
thinking out loud ...

simulating quatum systems
holographic projections based on viewable area
random number generators
factoring almost anything
large scale optimizations
database querying
ask a physist what he would do with it
any atomic level analysis
protien folding
drup interactions
nueral networking with quantum nodes
antialias_physorg
not rated yet May 07, 2012
I'd agree with most of those, though I'm not sure where AI/neural networking would benefit from quantum computing.

Conscience seems to be more a phenomenon of temporal coherence of (classical) parallel processing rather than superposition of states.

But anything where you have to sieve through large data structures could benefit enormously.
vacuum-mechanics
not rated yet May 08, 2012
By the way, it is interesting to note that nowadays we are so familiar with electromagnetism, so it is not surprise about our successfulness of conventional electronic computer.

In contrast to quantum mechanics in which we still do not know about its mechanism! But people are dreaming to get a quantum computer while we are not clear about quantum entanglement. So understanding the mechanism of quantum mechanics (in paper below), may pave the way to our dream!

http://www.vacuum...id=19=en