杏吧原创

Quantum computer users may soon have to wrestle with their own version of the 鈥淧C or Mac?鈥� question. A design based on superconducting electrical circuits has now performed two benchmark feats, suggesting it will be a serious competitor to rival setups using photons or ions.

鈥淭he number of runners in the race has just gone up to three,鈥� says of the University of Queensland, Australia, who builds quantum computers based on photons and was not involved in the new result.

The defining feature of a quantum computer is that it uses quantum bits or qubits. Unlike ordinary bits, these can exist in multiple states at once, known as a superposition. They can also be entangled with each other, so their quantum states are linked, allowing them to be in a sort of 鈥渟uper鈥� superposition of quantum states.

This means quantum computers could perform multiple calculations simultaneously, making them much faster than ordinary computers at some tasks.

Previously, setups using photons or trapped ions as qubits have made the most headway in early calculations. Now of the University of California, Santa Barbara, and colleagues have boosted the computing power of a rival design, first demonstrated in 2003, that uses tiny, superconducting wires instead.

Wire loops

Mariantoni鈥檚 team used a chip embedded with micrometre-sized loops of wire made of a mixture of aluminium and rhenium. When these wires were cooled to within a whisker of absolute zero, they became superconducting, meaning their electrons coupled up as structures called 鈥渃ooper pairs鈥�.

The pairs in each wire were made to resonate as an ensemble. Because each ensemble could exist as a superposition of multiple different resonating states, they acted as qubits.

Mariantoni鈥檚 team entangled these qubit wires using a second type of wire, known as a bus, that snaked all around the chip. First they tuned this bus so that it took on some of the quantum information in one of the qubits. Then they transferred this information to further qubit wires, thus entangling the qubits.

Benchmark tests

The design made strides in solving calculations often used as benchmarks for testing quantum computers鈥� capabilities.

It ran a calculation known as the quantum Fourier transform, which is a central component of the most famous quantum algorithm, known as Shor鈥檚. If Shor鈥檚 were run on a system with enough qubits, it would allow huge numbers to be factorised quickly. That has not happened yet, but if it ever did, it would cause many current encryption systems to break down, since they rely on the fact that ordinary computers can鈥檛 do this.

The researchers also used entangled qubits to create a system known as a 鈥淭offoli OR phase gate鈥�, which is a critical step towards building codes that do quantum error correction. This required entangling three qubits 鈥� a first for superconducting quantum circuits. 鈥淕etting three bits to play well together is hard,鈥� says White.

Ordinary chips

The advances may seem like baby steps, since both Shor鈥檚 algorithm and the Toffoli gate have been realised with relatively low numbers of photons and trapped ions.

But the reason the new result is exciting is that it could be hard to scale up these systems, which tend to be delicate and require specialised equipment, while the superconducting system uses chips like an ordinary computer. 鈥淭he beautiful thing about a solid circuit is that it鈥檚 something you can write using lithographic technology,鈥� says White. 鈥淚t looks much easier than say ion traps or photonic approaches.鈥�

But future quantum computers might not come down to an either/or choice like that between a Mac and a PC. Instead, true to their quantum nature, they may be 鈥渟uperpositions鈥� of different designs. 鈥淚 don鈥檛 think anybody knows what the best architecture will be,鈥� says White. 鈥淧robably we will end up using hybrids of the various approaches.鈥�

Journal reference: