Language is poor indeed when it comes to certain things. How does one describe the state of mind of someone reading up on quantum theory? Baffled. Boggled. Stupefied. Dumbstruck. Confused. On the list goes-without capturing the essence of the feeling one gets when trying to grasp that ultimate theory of physics, which describes all of existence.

After all, what theory tells you that something can be a wave and a particle at the same time? That a cat in a chamber can be dead and alive at the same time-until the chamber is opened? That there is only one electron, which manifests itself as many electrons in the dance that is the cosmos?

As if that were not enough, we have the multiverse. You guessed it, "multiverse" is short for "multiple universes": not a new concept, but one that is shoved in our faces with renewed vigour when we talk of quantum computing. You see, several scientists believe there are a multitude of universes, each identical to the other until a bifurcation point happens. Such a point comes about when a decision is made. To illustrate, say you're in a restaurant, and the waiter asks you "Tea or coffee sir?" Now, you have a choice, and decide on either tea or coffee-this is a bifurcation point, where the universe splits in two. In one of them, you're sipping tea, and in the other, coffee. These universes are only slightly different from each other-until more bifurcation points come along, and so on. That means infinities upon infinities of universes, and more interestingly, you are actually a billionaire with a harem in some universe!

"Shadow Photons"

But what's the point of considering other universes when we can't interact with them? Ah, but we can. Let's look at a simple enough experiment: light passing through two slits in an opaque barrier and forming a pattern on a screen beyond that. (Figure 1 on the next page.) We'll leave out the technicalities; when there are four slits, the pattern changes to what's seen in Figure 2! The light, as you know, consists of photons. So what is blocking the photons such that the dark regions appear in Figure 2? It's photons from the other universes, and this is what is usually called interference. These "other" photons are absolutely undetectable in "this" universe, and show themselves as existing only when they interfere with "real" photons. To repeat, that's the only way we know they exist, and that's the only way we can tell the existence of the other universes!

That is justification enough for the existence of other universes, for two reasons: one, this is a great way of explaining how photons interfere, and two, lack of space here doesn't permit further discussion!

So What?

We're getting there. We're trying to explain how "a quantum computer" is gazillions of times faster than a silicon computer. The concept of the multiverse comes in here, according to David Deutsch, who laid the theoretical foundations of quantum computing.

An aside: we must mention that not everyone believes there are gazillions of universes out there. But we're subscribing to that view for two reasons-one, Deutsch insists upon it, and we must respect that insistence; and two, there really does seem no other way Shor's Algorithm (up soon) would work.

Now, think of the problem of factorising large numbers. When you have two numbers such as 58359597 and 44845355, their product is 2617156845121935-easy enough to calculate. But when you have the latter number, how do you calculate what numbers it came from? In fact, factorising a 250-digit number could take longer than the known age of the universe on a regular computer! But Shor's algorithm-which is, well, an algorithm, devised by a man called, well, Peter Shor-actually does it in a reasonable amount of time. This, as it turns out, uses 10500 times the number of computational resources that seem to be present.

Deutsch asks, how can this have been achieved, when there are only 1080 atoms in the known universe?

The answer lies in "interfering universes." 10500 universes collaborate in the production of the answer, each of them doing a small part of the job. It's not like the problem is broken up into parts and given to each universe-there's no God out there to do that-it's just that this process is like the interference of light we talked about earlier. There, photons from other universes interfered with "our" photons, undetectable except for the fact that they interfered. Here, it's a mish-mash of answers from various universes that are interfering, leaving only one.

Now this might sound like rambling, and this writer might seem to be off his rocker, but that's common when dealing with quantum computing.

At this point you're probably screaming, "What is quantum computing?" Or "Just give me a description of a Quantel Inside processor!"

Sorry. We can't. That's just the way it is!

OK, we can shed some light on what's going on in the labs. For example, researchers have demonstrated cryptography working on quantum principles. Let's go through some real-world stuff.

Quantum Vagueness

Now we're getting into what a quantum computing environment looks like, what goes on, and how we get the answers.

The building blocks of a quantum computer are qubits instead of bits. A qubit can be not just one or zero, but can be both one and zero at the same time. So, if you have 10 bits, you can have only one of 210 values; if you have 10 qubits, you can have all 210 values at the same time. The qubit is based on the central ambiguity inherent in quantum mechanics: a property (such as spin) of a particle (such as an electron) is ambiguous until a process of disambiguation-such as an observation-causes the particle to "decide" upon what properties it has. Electron spin is too complex to go into here; just think of it as a property!

What does a qubit look like? Well, what does a bit look like? It's a voltage in a transistor, in most cases! And if that isn't hard to visualise, neither should a qubit be hard to visualise-qubits are represented by a quantum mechanical property of a particle such as an electron. They need to be set up such that they are in both states of spin at the same time. When one observes the state of the electron, the ambiguity vanishes, and this is called quantum decoherence.

An aside: we referred to a cat being both dead and alive until observed. That is one of the most popular paradoxes in quantum mechanics-what is the cat like until it is observed, and why should our observation make a difference? This is called the SchrÃ¶dinger's Cat paradox, and if you've been reading keenly thus far, you should straightaway get yourselves a copy of In Search Of SchrÃ¶dinger's Cat, by eminent physicist John Gribbin.

Now comes the meat of the matter: feeding in the input and getting the results! We need to present a series of qubits with a problem, and also a way to test the answer. This is done by setting up the quantum decoherence of the qubits such that only an answer that passes the test "survives" the decoherence. The failing answers cancel each other out "in a quantum way." Vague enough?

The series of qubits represents all possible solutions. This is simple enough to understand: a single qubit represents two possible solutions, and a hundred linked qubits mean 2100 possible solutions. The problem itself is formulated as a test to be applied to the potential answers. It is presented to the string of qubits so that they decohere, that is, "collapse" from their ambiguous states into real 1s and 0s that pass the test.

Now that's tough to digest, we know. We need an analogy. Ray Kurzweil, inventor, author, futurist, and strong-AI proponent, provides exactly what we need: when light strikes a mirror at an angle, it bounces off in the opposite direction at the same angle to the surface. But according to quantum theory, each photon actually bounces off every possible point on the mirror, thus trying out every possible path. The vast majority of these paths cancel each other out (in a quantum-interference way-now don't ask us how!), leaving only the "correct" path.

Now, in an abstract way, think of the mirror itself as the problem to be solved: "what is it that makes the light bounce off at this angle?" Only the correct solution-the light bounced off at the expected angle-survives all the quantum-interference cancellations. This is what we were talking about above-the test of the correctness of the answer is set up in such a way that most of the answers cancel out, and we're left with the one correct solution.

You've probably gathered that quantum computing is all about massive parallelism-and we mean massive. We don't mean massive as in supercomputer-massive-we mean it as in giga-massive. A quantum computer doesn't just work "much faster" than a regular computer-it ropes in billions of our friendly neighbourhood universes to do in a matter of minutes what your rig would literally take aeons to do.

But then, that's not all good. It means we'll probably never have a general-purpose quantum computer-it'll only be good for tasks that require massive parallelism.

Decoherence Revisited

We've spoken about how decoherence is the process of removing the state ambiguity of a particle. But there is a more sinister face to the phenomenon: it prevents easy construction of a large-scale quantum computer.

The quantum theory of parallel universes... is not some optional interpretation emerging from arcane theoretical considerations. It is the only tenable explanation of a remarkable and counter-intuitive reality"

David Deutsch, Physicist,Oxford University

So OK, how do we build one? We start with simple quantum logic gates, and integrate them into quantum circuits, of course! Easier said than done: gates and circuits are familiar enough at our level of size, but they're much more complex at the sizes where quantum effects come in. Still, we do already have quantum logic gates, simple devices that perform one elementary quantum operation, usually on two qubits. In fact, the first quantum logic gate was built using light in 2003. They differ from regular logic gates in that they perform operations on quantum superpositions (the ambiguous states of particles).

Now here are the practical problems: as we increase the size and complexity of the circuits we're merrily imagining, the number of interacting qubits increases, and this makes it harder to design the interactions that would exhibit quantum interference! (If you've been following, interference is essential.) One of the biggest impediments to progress is that the surrounding environment should not be affected by the interactions that give rise to quantum superpositions.

It's something like a process going on in a closed room; think of it as secrets being discussed inside. When the circuits get more complex, the "secrets" will spill to outside the room. Formally, this "spilling" of information to the outside is also referred to as decoherence. Now this is a totally different way of looking at the phenomenon. Don't ask us how the same thing can be thought of in two such different ways-that's the way it is!

To clear the fog for some readers, there is disagreement within the community-and different beliefs-about how to define decoherence. Different writers will tell you different things, and let's just leave it at this: if a particle can be in two states at once, why not a concept?

Breaking News

In October 2004, a quantum memory register using caesium atoms was built at the University of Bonn. Researchers were able to "write" to the register using microwave radiation, which puts the electrons into a position between their two natural orbits around the nucleus.

In July 2006, US NIST researchers brought quantum computers closer to mass production. The development reported was that of the construction of a two-dimensional ion trap; ion traps have proved to be the best way to make qubits, allowing up to eight to be connected together. (Kurzweil estimates that at 40 qubits, quantum computers will outdo todays PCs.)

In August 2006, a team of researchers at the Delft University of Technology in The Netherlands created a device that could manipulate a single electron using conventional chip fabrication technology. A major benefit of making a qubit this way is that the qubits will be easier to scale up.

So, then, we've got the memory, we've got the processors, we've got scalability. At this rate, we have good reason to believe we'll see a working quantum computer in our lifetimes!

Concluding Remarks

Quantum computing is the future. Repeat that a hundred times. But why? Well, for the simple reason that it's quantum mechanics that really governs the universe (or multiverse, if you will).

Our relatively feeble attempts at computing thus far have been limited to the application of rather elementary laws of physics. Even such things as DNA computing don't really exploit truly fundamental laws. After all, biology and chemistry are physics.

With quantum computing, we've reached the final frontier-the exploitation of the most successful theory of reality. We're admitting and harnessing the existence of parallel universes: we're taking hold of the very fabric of reality.