As demands for processing power continue to increase, quantum computing is often cited as the next critical step in expanding our technological capability. But just how does quantum information processing work? Will we have to change our entire approach to software and hardware design? And how soon is it coming?
Before 1989, the word "quantum" was rarely heard anywhere outside the relatively small global community of theoretical physicists, who generally kept their quarks and their electron microscopes to themselves. The word itself originates from the Latin quantus, meaning "how much", and had fallen into disuse as people realised there were better things to do with their time than mess around with classical obscurities.
Nonetheless, in the past two decades, a pair of developments have resulted in the phrase "quantum" (followed by the noun of your choice) popping up much more frequently in general conversation. The first--and arguably the most visible--was the TV series Quantum Leap, which ran from 1989 to 1993 and starred Scott Bakula as a time-travelling scientist who managed to get into many exciting scrapes which could nonetheless easily be resolved in an hour (commercials included).
The second--and undoubtedly the more significant from our point of view--has been the emergence of quantum computing as a potential pathway to the information future. In any discussion of how technology is set to evolve in the twenty-first century, quantum computing will invariably get a mention. But what exactly is it? Can we make use of it right now? And do we have a hope in hell of really understanding it without an advanced degree in physics and a solid understanding of complex mathematical theories?
Principles of quantum computing
While the correct answer to that last question is probably "no", the fundamental ideas behind quantum computing are fairly well understood, and were being discussed back when computer systems were still the size of two-bedroom apartments. Quantum theory in physics refers (allowing for a necessary amount of massive oversimplification) to the idea that energy from electrons is not discharged continuously, but in discrete quantities. These are referred to as quanta (which, for the linguists amongst you, is the correct plural form of quantum).
The concept of quantum computing arises from the nature of quanta: rather than simply representing two alternate states (as is done in classical binary notation), quantum information can be represented as superpositions of those states. Indeed, quantum physics holds that when interacting, all possible states that can be realised are considered, rather than just those which eventually occur.
This phenomenon is represented in quantum information theory by probabilistic bits (known informally as pbits). A probabilistic bit represents (for instance) the not yet determined outcome of a coin toss. Before the coin is tossed, it could equally fall as heads or tails. Once it has fallen, it can be represented in classic binary form. Pbits provide a means of translating from quantum information to the more familiar forms we use (often referred to by quantum theorists as "classical" information.)
A useful (if somewhat lazy) way of thinking of it is this: bits can be represented as two distinct dots, pbits as a single line connecting the two dots, and qubits (quantum bits) as a series of lines which might potentially connect the two dots, possibly in the form of a sphere for imaginative convenience. This complexity makes them harder to comprehend (which is why we've glossed over lots of the details here) and much harder to devise processing systems for, but rather more powerful when those problems are overcome.
Given that most of us have perfectly functional laptop PCs already, why would we want to do any of this? The short answer is that certain kinds of mathematical problems have proven very difficult to solve using conventional techniques, but appear to be rather more easily handled within quantum computation systems. Complex calculations are essential to all kinds of everyday computing activities, from generating three-dimensional images to creating sophisticated encryption systems.
A relatively easy-to-understand example is that of prime numbers (numbers which don't have any whole number divisors other than themselves and one), and factorisation (breaking a number into its component numbers) generally. Given any individual number, we can work out if it is prime simply by checking each number smaller than its square root and seeing if it is a proper factor of the original number. If any are, then the number isn't prime.
While relatively easy to understand and implement, this process becomes increasingly time consuming as the numbers get larger (even calculating the square root itself becomes more complex). It would be appealing if we could simply apply a single test or formula to find the factors of a given number and hence determine its primeness or otherwise. However, to date there is no general formula which allows us to determine if a given number is prime simply by examining it (though there have been some tantalising attempts). Quantum computing holds out the promise of making such calculations easier, as it provides entirely different means of factorisation.
This has immediate relevance in the field of cryptography, which relies on using numbers that are relatively difficult to factorise to ensure that keys remain secure. If a fully functioning quantum computer could be built, current cryptographic protocols would become much easier to crack.
At the same time that quantum theory has been evolving, conventional computer technology has been progressively shrinking, to the point where information in binary form is being transmitted using smaller and smaller circuitry. Eventually, it is expected that information will be stored in essentially atomic form--and at that stage, quantum physics and conventional computing will inevitably collide.
"Quantum computing begins where Moore's Law ends--about the year 2020, when circuit features are predicted to be the size of atoms and molecules," says Isaac L Chuang, a former IBM researcher in the field of quantum computing and academic at MIT. Essentially, two paths are available at this point: trying to make atoms behave like traditional computers, or taking advantage of the quantum behaviour of atoms to create new computing models.
One immediate problem with quantum computing is that the behaviour of atoms is not easily controlled. Ensuring that quantum computers maintain a predictable state is one of the biggest current challenges. To date, most proposals have required extensive error correction systems, which tends to diminish the raw power theoretically available from a quantum system.
The concept of quantum computing is not a new one. Models for quantum computing systems were first described in the 1970s, but at the time many scientists doubted that practical quantum computing systems could ever be constructed. A critical breakthrough came in 1994, when AT&T research scientist Peter Shor demonstrated that quantum calculation systems could be used for factoring large numbers much faster than existing systems. (This was purely a theoretical demonstration; a computing system using Shor's algorithm didn't appear until 2001, as we'll discover below, and its power was not exactly overwhelming.)
Since that time, a number of experimental quantum computing systems have been constructed, most often by research teams drawing together experts from academia and commercial computing suppliers. It also remains a subject of interest to the broader physics community, since quantum computers are expected to be a useful tool in simulating the behaviour of other quantum-influenced phenomena.
Who's doing it?
All researchers in the field agree that full-scale commercial implementations of quantum computing are still some way off. You're not likely to be buying a desktop system with a quantum processor in the next few years; more pessimistic researchers don't believe you'll be buying one in your lifetime. Nonetheless, many existing IT companies are carrying out extensive research in the quantum computing field, and hoping to apply at least some of its principles to existing applications.
IBM has been deeply involved in quantum computing research for some years. In 2000, the company's Almaden Research Centre, under the direction of Isaac Chuang, demonstrated one of the first practical quantum computing systems, which used five qubits made from fluorine atoms. Using this system, the research team was able to simply carry out order-finding for a given function--the kind of mathematical problem which is relatively straightforward with quantum computing systems, but extremely difficult using conventional binary processors.
At the time, IBM explained the order-finding problem as follows: imagine a building with a large number of rooms and an equal number of randomly placed one-way passages. Some of the passages connect rooms, while others loop back onto the room they started from. Anyone moving through all the rooms and passages will at some point return to the starting room, but what is the minimum number of passages they will pass through before doing so? The IBM system was able to solve any version of this problem, involving any combination of rooms and passages, with just one step; conventional mathematical systems would need up to four steps, depending on the complexity of the rooms and passages involved.
A team working under Chuang showed further progress in 2001, when a seven-qubit system was constructed which was capable of implementing Shor's factorisation algorithm, which had only existed as a theoretical model up until that point. "This result reinforces the growing realisation that quantum computers may someday be able to solve problems that are so complex that even the most powerful supercomputers working for millions of years can't calculate the answers," Nabil Amer, manager and strategist of IBM Research's physics of information group, commented at the time.
The word "someday" is used very advisedly. While mathematically interesting, such problems don't have immediate application in the world of practical computing, and progress since that time has, understandably, been slow. Chuang predicts that a quantum computer will need to be able to manipulate several dozen qubits, if not thousands, in order to be practically useful. No-one has yet come close to being able to control even small groups of atoms in an ongoing practical way.
Microsoft has also invested heavily in quantum computing research, but, like IBM, has some reservations. Christian Borgs, a senior researcher with Microsoft Research's theory group, argues that the extensive error correction required to ensure that quantum computing systems function effectively (because of the interference between separate qubits) may not make them commercially viable, even if they are theoretically interesting.
| ||||||
 | ||||||
As Microsoft's remarks make clear, the development of quantum computing has proceeded hand in hand with evolutions in nanotechnology. Nanotechnology involves the creation of machines at the atomic level, manipulating objects atom by atom to build machines which can themselves then carry out further atomic building activities. (For a detailed overview of nanotechnology and how it is being put to practical use today, see Practical Nanotechnology).
While it has had plenty of problems of its own, so far nanotechnology is a clear winner in the practical-application stakes. Nanotechnological systems are already been used for everything from creating processor designs and storage systems to manufacturing clay for car parts and sunscreen.
By way of contrast, the most visible usage of quantum computing has been in quantum cryptography--and this doesn't really qualify as quantum computing in the strict sense (see Stick to the encrypt below for details of how it works and why people get confused).
One potential delay to developments in quantum computing has been the continuing massive growth in processing power, and the parallel potential which emerges from technologies such as grid computing, which let thousands of relatively cheap systems be lashed together for supercomputing power. If this power can be harnessed cheaply enough, the necessity of creating a new methodology seems rather less urgent.
To return to our prime numbers example: while it isn't a trivial task to work out the prime factors of a given large number, it does get easier every year as computing power increases. Sure, the methodology may be primitive, but it works.
One reason for the recent boost in prime searches has been projects such as the Great Internet Mersenne Prime Search, a system which allows users to search for primes using spare processor cycles, and share the results via the Internet. (Mersenne primes, if you were wondering, are numbers that can be expressed in the form 2 to the power of something minus 1.) The methods used for determining these are mathematically fairly unsophisticated, but the sheer volume of calculations being carried out still makes them useful.
Computerised systems have already been used to discover prime numbers with more than four million digits, and many people expect that a 10 million digit prime will emerge soon. By way of contrast, the theoretical quantum model constructed by IBM in 2001 only discovered the factors of 15 (five and three). This proved Shor's theory to be valid, but hardly told us anything we didn't already know, other than confirming that building a quantum computer is still mind-bogglingly difficult.
Some scientists take the view that all natural systems fundamentally work in the way suggested by the classical method--combining extremely simple elements and operations in huge quantities to create distinct and complex environments. Given that position, development of quantum information systems might be unnecessary, and even counter-productive. However, since quantum theory is also widely accepted as a key scientific principle (and can also be said to work in this basic way on some level), this is not an argument which is likely to be decided any time in the near future. n
Stick to the encrypt
One of the most widely discussed concepts in quantum computing is quantum cryptography. Strictly speaking, this encompasses two separate problems: whether quantum computers can be used to crack existing cryptographic systems, and whether quantum computing will provide new methods for secure data exchange that are fundamentally different to existing models.
As we've discussed elsewhere in this article, quantum computing holds up the promise of quickly identifying the given factors of any large number. This in turn threatens the stability of existing cryptographic systems, which use very large numbers as cryptographic keys, and rely on the fact that it's currently very difficult to find the factors of these large numbers to provide security.
While this is a theoretical concern, is it actually going to happen any time soon? Probably not. Researchers estimate that breaking a 1000-bit key would require a system using more than 3000 qubits--several orders of magnitude ahead of where current systems are. People with a Pentium and time on their hands remain a more immediate threat (and for many purposes even they can be ignored).
What is normally described as quantum cryptography is actually a combination of quantum and classical systems, exploiting the properties of quantum mechanics to enhance the system of key exchange that is the backbone of most modern cryptography.
Such quantum cryptography systems exchange secure keys using a quantum communication channel (photons transmitted via optical fibre), as well as an authenticated classical channel for encrypted data.
This limits the transmission distance, but provides a means of creating a key which can't be intercepted without detection (because of the way in which photons randomly change their states when read without an agreed basis). So far, such systems also remain largely experimental, but they hold more potential for future development that quantum computing as such at this stage.
Executive summary
Quantum computing makes use of the physical properties of atoms (outlined in quantum theory) to create methods of computing which are fundamentally different to conventional computers, and potentially much more powerful. This path is being pursued partly because of its inherent interest, and partly because it is widely expected that as processor components continue to get smaller, they will begin operating on the atomic or sub-atomic level anyway. The power of quantum computers is measured in qubits; to date, a system using even 10 qubits has proved impossible to construct.
- What will be the applications when they do build one? Quantum computing is currently mainly of interest for solving complex mathematical problems which are difficult to handle with conventional computing systems. These may not appear to have immediate applications, but some may prove useful in the long term. For instance, work on number factorisation in quantum computing may have a significant impact on the future development of cryptography.
- Is it going to change my life? In the next decade, probably not. While the underlying theories of quantum computing are well understood, no practical devices have yet been realised, in part because individual control of atoms remains so difficult. It's also not yet clear whether quantum computing systems will prove more economical to produce, even if they are in an a theoretical sense more efficient. Nonetheless, considerable efforts are being invested into developing this area further.
Subscribe now to Australian Technology & Business magazine.
7%
2%
Hello,
The whole of quantum mechanics can not only be aided by complex mathematics but clarified to form a Unified Field Theory.
I am explaining my new complex quantum mechanics to show my new theories and how I have rewritten complex mathematics at my new yahoo site.
In fact as a mathematician who can resist knowing the theories below?
Mathematically I hope that you can appreciate that to be serious about rewriting complex mathematics you need to overcome 2 insurmountable problems (you may need to scroll down at yahoo to the relevant messages) that include:
1> The interpretation of how i^2 = -1 (necessary to base any new revision of complex mathematics.
2> How to order the complex set of numbers.
3> There is also an explanation on How you can use this to solve Newton's Cube (the cubic nature if light's expansion).
Please feel free to advertise or debate the material on site elsewhere.
A written letter is always responded to and appreciated.
Thanks for your time and support,
From:
Mr Alexander Ross
7, Midland Street
Accrington
Lancashire
United Kingdom
BB5 2AX
http://tech.groups.yahoo.com/group/complexquantummechanics