“You see a lot of attention to quantum hardware, and few Quantum software. But the latter, you need to harness the power of quantum computers. And making quantum software is a lot harder than traditional software, “said Harry Buhrman, computer scientist at the Center for Mathematics and Computer Science (CWI) and the University of Amsterdam (UvA). He wants QuSoft exploit this largely undeveloped land.
Harry Buhrman, during an episode of The University of the Netherlands.
Water
QuSoft is the software equivalent of QuTech This is a conglomerate of physicists from TU Delft, where a company like Microsoft is pumping in millions. Thus, large-scale and financially QuSoft will not. Yet this partnership CWI, VU University and the University of Amsterdam, will now fifteen investigators, grow to 25 researchers later possible to forty.
In recent years the development of the quantum computer momentum hit. More and more components of which are still hypothetical, super fast computer are realized, and it will succeed always better to let them work together. Some experts dare to say that we over a decade, the first working quantum computer, and it may be that the then state in Delft. It seems really possible to build such a device leads to great enthusiasm among the physicists involved. But many of them forget for convenience, there also need to be resolved still difficult software issues.
This and illustrations below show impressions of a Bose-Einstein condensate, an exotic state of matter in which all of the atoms at ultra-low temperature come up with one another in so-called “superposition.” A quantum computer is like a Bose-Einstein Condensate a macroscopic quantum object, an object with ‘everyday’ dimensions which still has the strangeness of the quantum world.
Quantum Software
The way in which charges a quantum computer differs fundamentally from that of a conventional computer. Bluntly you could say that the software of a classical computer every job breaks down into a series of simple operations on a large number of individual bits – which at the beginning each have the value 1 or 0 – and the computer sequentially performs.
a quantum computer, however, can already with a relatively small number of qubits (eg, fifty) function, which together 2 50 , or about 1,000,000,000,000,000 common conditions ( ‘ superpositions) can be. The quantum software for a certain job is made up of a series of quantum-operations that are performed on single or pairs of qubits.
Physical can you a qubit proposals as, for example, one isolated electron, which can exist in two states ( spin up / spin down, similar to anti-clockwise / clockwise turning) or in a superposition of both. First, the initial state of all the qubits is brought into superposition, for example, with pulses radio radiation. Finally, a measurement is carried out on the whole system, so that it ends up in the state that corresponds with the correct answer.
Encryption
Because the number of possible superpositions doubles with each additional qubit, to say that the computing power of a quantum computer grows exponentially with the number of qubits. But it is much too simple to state generally that a quantum computer exponentially faster than a classical computer. Buhrman: “Not every problem late accelerate exponentially. You should now investigate quantum software can and can not, because if the quantum computer is, it is too late. “
A good example is the algorithm of Shor. That’s one of the first quantum algorithms, the prime factors of a very large number exponentially faster then calculates the fastest now known classical algorithm. A small number, such as 182, dissolve into prime factors is not so difficult. Because you just try to share with the smallest prime number, 2, and then by the next prime numbers 3,5,7,11,13,17 …. and then you find that 182 × 7 × 2 = 13. But it is impossible in practice to very large numbers (of two hundred or more digits), thus finding the prime factors. And that is exactly where the security of the RSA cryptography is based on.
RSA is a very special secret because it is asymmetrical. Each user has a personal, public key (based on a very large number N), which knows all, and that anyone who can send individual encrypted messages. Only the user also has the secret key (based on the X prime factors) that allows you to decode these messages. A large part of the security of Internet payment systems, and makes use of RSA.
Buhrman recalls his colleague Gilles Brassard, who warned years ago: set the quantumcompter there, and you only came to the discovery that it lets you crack RSA encryption. That was a huge disaster.
Cryptographers warn that those who now protect information that must remain secure in ten years, should not use more of RSA. Because you may assume that the US National Security Agency and other intelligence agencies many of which RSA encrypted data already in the store are awaiting the arrival of the quantum computer.
Quantum Authentication
The research program of QuSoft has four legs. The first leg is called “Applications few qubits. Before the first full-fledged quantum computer is there, there will be experimented with prototypes with a few qubits. There is clear: the more qubits, the more difficult to bring them in superposition and keep. Basically programming computer scientists in terms of universal gates. These are the most basic operations that a computer performs. “Why does an algorithm in many different platforms and types of computers. But as physicists in the lab only nine qubits have available, then program very platform-dependent.
Still, you want a simple, with those few qubits perform quantum algorithm. we can develop together with the people in the lab. Ten qubits provide a superposition of 2 10 , or about a thousand possible states. Then you can even calculate what the system does. About a few years, with fifty qubits and superposition of 2 50 situations, this is even a classical supercomputer no longer count on. “
Rommel
the second leg of QuSoft engages in “Cryptography in the quantum world.” It is already known that the RSA cryptography is not “quantum resistant. What other ciphers are not either, and what is it? For another widely used cryptography, AES, no quantum algorithm is known to crack it accelerates exponentially. Probably not even exist, but how sure are we that? And how do you prove that? For rapid, classic encryption which is nevertheless quantum resistant, will continue to exist a great need. “An important aspect of this leg is to investigate what kind of new cryptographic ideas we can use quantum mechanical effects. An example is localized cryptography, which we are developing. “
The third leg of QuSoft looking for ways to eliminate errors in quantum algorithms and verify that a quantum computer actually does what he has to do. This “quantum verification ‘is far from simple. Classic software you step by step be carried out by the computer, so you can see exactly what happens. That is a common method to extract defects from software. “But if you’re a quantum computer halfway through stop a program, indicates that just clutter.” The early readout of the state of the quantum system produces namely random data
The result is a curious problem:. You you may think you’ve figured out a clever quantum algorithm, which have been translated into proper software, but how to verify that? Buhrman: “That is out of control in the classical way.” How should you? Buhrman makes a comparison with ‘blind’ tests: “Suppose I think I very well the difference between Pepsi Cola and Coca Cola can taste. If I one hundred times out of two paper cups with coke to choose Pepsi, and subsequently that I was good it all a hundred times, it gives a lot of confidence that I can really taste the difference, even though I have no idea how I exactly Doc. “So you can quantum software as a black box consider you subdue blind tasting of this type, and if you often repeat enough, you’m pretty sure the software works correctly.
No consumer product
According to Buhrman this is part of a larger problem, which even touches on the limits of the scientific method. “How do we know if a quantum computer working properly? We can, because he check delivers an exponential acceleration of computational processes, not with a classical computer if the answer is correct. The bottom line is that we are no longer the model for quantum computer can calculate, and so can not verify the answers he gives, knock. “
Finally, there is the paw ‘quantum architecture.” This is dot on the horizon a quantum programming as easy to use as Pascal or Python programming languages with which every enthusiast now own computer programs can write that requires many years development Buhrman:.. “We are now at the stage where was the classic computer in the forties, when program was a matter of plugging in connectors and wires soldering. “Incidentally expect even the most enthusiastic quantum physicist or -informaticus not that the quantum computer will be a consumer product. It will be a very expensive and complicated device, which delivers in a specialized center quantum computing to internal and external users such as classical supercomputers do.
No comments:
Post a Comment