Home / Microsoft / A New world record is made in quantum simulation

A New world record is made in quantum simulation

Australian physicists achieved the most important quantum computing simulation so far, in computer science.

If we consider quantum computer a racing car it would not only go faster than a Formula One, but it would simply take a shortcut to appear on the finish line just after the starting line. And if you came to look under the hood to see how it works, the engine would quickly decompose into a single random element, like a spark plug.

This is the strangeness of the quantum world in which the normal laws of atomic-level physics become, as Einstein said, “rare.”

A quantum computer takes advantage of quantum physics to quickly find the right answer to a problem by carefully analyzing the probabilities and adjusting them, while a classical computer will consume time and memory by analyzing each possible response one at a time.

But physicists at the University of Melbourne have shown that classic computers still have a lot of life ahead of them. Scientists have set a new world record in the simulation of quantum power in a classical computer, showing that it has more capacity to perform the monotonous work of processing quantum data than any of the current prototypes of quantum computers on a small scale.

The simulation of the quantum to understand it

This means that scientists have a new and powerful tool to grasp and understand the quantum state and develop quantum software. Ultimately, it will help us understand and test the types of problems for which a possibly larger quantum computer will be employed as the quantum hardware develops over the next decade or so.

“The ability to simulate quantum algorithms at this level is important to learn how a quantum computer will work physically, how the software can work and what kind of problems it can solve,” explains Professor Lloyd Hollenberg, head of the University’s Thomas Baker Chair. from Melbourne, who directs the team and is deputy director of the Center for Quantum Computing and Communications Technology.

New world record in quantum simulation “We are at the dawn of a second technological revolution”
New world record in quantum simulation Polynomials against quantum computers
Nowadays, the prototypes of quantum computers are too small to do anything useful that a classic computer can not do anymore. But quantum hardware evolves rapidly, and it is likely that quantum computers are much more powerful than classical computers in solving some problems for two quantum oddities: the “superposition” and its even rarer cousin, the “entanglement”.

A matter of focus

Classic computers work with programming bits, the most basic form of data. The bits are binary, that is, they are 0 or 1, and they are programmed to encode and process data. But in a quantum computer, bits, or cubits, are quantum mechanical objects like atoms. Quantum states can also be binary and can be placed in one of two possibilities, or in both at the same time. The quantum superposition means that two cubits can be, in a certain sense, the four combinations of 0 and 1 at the same time.

That unique ability to process data is further enhanced by entanglement, in which the state of a cube when measured mysteriously determines the state of another cube.

A representation of quantum computing in action showing the “forest” of different probabilities that the machine uses to guide it more effectively towards the response to a problem. The example above is a simulation of a quantum computer that finds the prime factors of a number using the Shor Algorithm. Photo: Matthew Davis, Gregory White and Aidan Dang

The simulation of cubits and their quantum processes, or “programs”, in a classical computer are a fundamental step to understand how a larger and more useful quantum computer will ultimately work.

 New world record in quantum simulation

The problem is that the use of conventional techniques to simulate a random quantum process that is significantly larger than any of the current quantum prototypes, would soon require what Professor Hollenberg describes as “planetary scale” memory in a classical computer.

To overcome this obstacle, his team gave the simulation a specific mathematical problem to solve it. Being specific, they did not need to simulate the entire quantum state to simulate quantum computing on a larger scale in action.

Imagine 1,000 million laptops
To get an idea of ​​the enormous memory capacity of quantum computing, one of the largest prototypes, IBM’s new 50-cube machine, could in principle simultaneously represent approximately one billion trillion combinations of numbers.

To simulate a random quantum state, the machine would use about 18 petabytes of classic computer memory, or the equivalent of more than one million laptops with a RAM of 16 gigabytes. The IBM researchers have been able to simulate in a classical way so far 56 cubits in carefully chosen states.

But the Hollenberg team has gone much further and simulated the performance of a 60-cubic-inch machine, which would have required about 18,000 petabytes, or more than 1,000 million laptops – much more than the largest supercomputer – to represent the entire quantum space of numbers.

“A really random state of about 50 cubits is pretty much the limit that can be simulated today, but if you think of a quantum computer doing something useful like running an algorithm, it’s no longer in a random quantum state, but in a specific one whose Simulation may require much less memory, “notes Professor Hollenberg.

The Magnus supercomputer at the Pawsey Supercomputing Center in Western Australia that researchers use to perform their quantum computing simulation. Photo: Pawsey Supercomputing Center (used with permission)

Aidan Dang, a graduate in Science from the University of Melbourne, has made and developed the simulation that has broken the record by asking him to find the two prime numbers that when multiplied are equal to the semi-prime number 961.307. A quantum computer would do the job using 60 cubits that used the Quantum Factoring Algorithm formulated by the mathematician Peter Shor. This calculation of the factorization can be done on a laptop, but it exceeds the limit of what the prototypes of current quantum computers can solve. However, the Melbourne team simulation was able to solve it like a quantum computer with 60 cubits using only 13.8 terabytes of memory in Pawsey’s classic supercomputer in Western Australia.

“We almost used up all the time they gave us for the simulation at the Pawsey Supercomputing Center, but we did it,” says Dang. “Now we can use the results to get clues about how the first full-scale quantum computers will work.”

Decoders of supercodes

The difficulty of factoring semi-prime numbers is key to Internet security, because when using large numbers with many digits it is almost impossible for classical computers to calculate the factors to decipher the security key.



Leave a Reply

Your email address will not be published. Required fields are marked *

KI Media Tech 2019