|
|
|
|
from TheEconomist Website
QUANTUM mechanics and computers traditionally don't mix.
The strange fuzziness of the quantum world is a big obstacle for chip designers, who work with components so small that quantum effects make the electrons flowing through them unruly and unpredictable. But it is possible to design a computer in which that quantum fuzziness is a feature, not a bug.
Researchers have been working on so-called quantum computers since the early 1980s, when the idea was first proposed.
Recently, a Canadian firm called D-Wave has been in the news, for its device - a special kind of quantum computer designed to solve one particular problem - has, for the first time, been raced against a classical, non-quantum computer to see which is faster.
D-Wave's machine is designed to solve only a specific kind of problem, but scientists around the world are working on general-purpose quantum machines that could attack any kind of problem that a standard computer could tackle. But what exactly is a quantum computer?
Classical computers - like the one on which you are reading this article - work by performing a series of simple tasks (such as adding two numbers) extremely quickly.
But the circuits in a classical computer abide by the rather boring laws of classical physics, which stipulate that they can only be in a single state at a given time. Quantum computers use the racier laws governing quantum mechanics to skirt around that limitation.
The fundamental unit of quantum computation is the "qubit", the quantum analogue of the ordinary "bit" in a standard machine.
Like ordinary bits, qubits can take the value of 1 or 0. Unlike ordinary bits, their quantum nature also lets them exist in a strange mixture - a "superposition", in the jargon - of both states at once, much like Erwin Schrödinger's famous cat.
That means that a quantum computer can be in many states simultaneously, which in turn means that it can, in some sense, perform many different calculations at the same time. To be precise, a quantum computer with four qubits could be in 24 (ie, 16) different states at a time.
As you add qubits, the number of possible states rises exponentially. A 16-bit quantum machine can be in 216, or 65,536, states at once, while a 128-qubit device could occupy 3.4 x 1038 different configurations, a colossal number which, if written out in longhand, would have 39 digits.
Having been put into a delicate quantum state, a quantum computer can thus examine billions of possible answers simultaneously. (One way of thinking about this is that the machine has co-operated with versions of itself in parallel universes.)
With great power, alas, come irritating limitations.
The answers that a quantum machine gives to questions are probabilistic. In other words, they might be wrong and must be checked. If a given solution is wrong, the calculation must be repeated until the correct answer emerges, a flaw that removes the speed advantage quantum computers offer over classical devices.
Clever programming can exploit another quantum phenomenon called interference to significantly improve the odds of getting a correct result, restoring some of the speed-up. But boffins have figured out how to do that only for a small number of problems, which limits the probable utility of quantum computing.
Most famously, by running something called Shor's algorithm, a quantum computer could rapidly find the prime factors of very large numbers, a computational feat whose difficulty for ordinary computers forms the basis of many of the cryptographic systems used on the Internet.
That would certainly be useful.
But for many tasks, ordinary computers will probably be just as fast as their quantum counterparts.
Powerful Quantum Computing Technology from PandoDaily Website
As the NSA snoopapalooza continues into its second week, much has been made of its vaunted Utah Data Center, a billion-dollar facility spanning 1.5 million square feet that can process data on the order of zettabytes, or 1021 bytes. (That’s 10 to the 21st power. If you know how exponents work, that’s a hell of a lot of bytes).
While storing data can be burdensome enough - one estimate says the NSA will pay $40 million a year on electricity bills alone - analyzing it and figuring out what’s relevant is even harder.
As “The Wire” creator David Simon wrote in a thought-provoking though overly-simplified piece on NSA snooping,
There’s just too much data to do anything useful with, he says.
Obviously there’s still a lot we don’t know about the scope and capabilities of the NSA’s surveillance program (they don’t call them “secret agents” for nothing). But generally speaking, it is true that after a certain point, the more data you collect, the more additional data you need to gather meaningful insights from it.
But that’s if you use boring classical computers. What about quantum computers?
Unlike classical computers, which rely on bits of information that are either 0 or 1 (or on/off), quantum computers rely on qubits which exist in a quantum state, allowing them to be 0, 1, or any number of infinite superpositions in between.
This allows computations to be processed in parallel, and scientists say it’s theoretically possible to build quantum computers that are a million times faster than their classical counterparts. (For more, check out our interactive explainer).
Alas, we’re still in the research stage of the quantum computing revolution, and quantum devices have a long way to go before they could consistently outperform most commercial laptops.
But that hasn’t stopped organizations like Google, IBM, and, yes, the NSA, from investing in the technology.
And what might it do with this technology?
Read what Google has to say about the potential of quantum computing:
Sounds like a great way to find balls in drawers. It also sounds ideal for identifying individuals amidst a vast amount of data. If NSA surveillance is like finding a needle in a haystack, then quantum computing could be thought of as a giant magnet.
Besides surveillance the NSA has another good reason to invest in quantum computing: cryptography. One of the more promising applications is its potential to crack some of the toughest encryptions known to man, the kind of stuff used to protect state secrets and financial transactions.
If a key is developed by other nations, the NSA needs to be prepared to safeguard our own data.
But the NSA could also go on the offensive, both against foreign nations and potentially against its own citizens for the purpose of surveillance. Today, we’re worried about the NSA accessing our phone records and Facebook status updates.
What happens when they can access our bank statements or medical records?
True, they can already access this information with a subpoena or court order. But when an individual’s encrypted information can be mined, located, and then cross-referenced with a thousand data points - from what porn, er..., websites you visit to what books you have on your Kindle to who you call and text, and all at the speed of a million MacBook Pros - there’s enormous potential for abuse.
But how worried should we be?
According to MIT professor of electrical engineering and computer science Scott Aaronson, we can breathe easy, at least for now:
So which is it:
Aaronson compares it to the first computing revolution.
And yet, when Congress passed the Patriot Act in 2001, few would have predicted that 12 years later over half of American adults would carry tiny computers in their pockets capable of collecting a staggering amount of data.
We may be 10 years, 50 years, or 100 years from the age of the quantum computer, but when that happens, we need to prepare ourselves, both with new legal protections and encryption techniques.
Because a world where organizations can collect any piece of recorded data on an individual instantaneously may be just around the corner.
|
