Quantum Computing

Microsoft unveils first quantum chip powered by 'topological qubits'

Microsoft unveils first quantum chip powered by 'topological qubits'
Microsoft says its new chip represents a major breakthrough on the path towards real-world quantum computing
Microsoft says its new chip represents a major breakthrough on the path towards real-world quantum computing
View 2 Images
Microsoft says its new chip represents a major breakthrough on the path towards real-world quantum computing
1/2
Microsoft says its new chip represents a major breakthrough on the path towards real-world quantum computing
The Majorana 1 chip currently fits eight qubits, but can hold more than one million of them
2/2
The Majorana 1 chip currently fits eight qubits, but can hold more than one million of them

Microsoft says it's made a major breakthrough in quantum computing capabilities with the Majorana 1, its first quantum chip, and the first of its kind to be powered by what are called topological qubits.

What's this new chip all about?

First off, the palm-sized processor has eight qubits at its core. These qubits are made from a new class of materials called 'topoconductors,' and they're said to be awfully small – about 1/100th of a millimeter each. They're also fast and can be digitally controlled, which means a large number of them can be more easily managed than in other quantum computers.

Microsoft is also excited about the possibility of fitting up to a million qubits onto this processor. At that milestone, you could expect a quantum computer to operate reliably, i.e. with enough error correction, to begin tackling real problems that surpass the capabilities of today's classical computers.

That includes challenges like simulating complex molecules for drug development, optimizing chemical reactions for more efficient fertilizer production (which currently accounts for 5% of all greenhouse gas emissions worldwide), modeling new materials for better batteries or solar cells, and running complex financial calculations to address macroeconomic issues.

The processor isn't commercially available, of course, and will likely only be used for evaluation, simulations, and developing future chips.

Majorana 1 Explained: The Path to a Million Qubits

How did Microsoft make this?

Perhaps the most interesting part of Microsoft's announcement is about how it got here. The company says this is one of it's longest-running research projects, and it's been 17 years in the making – with an even longer history of theoretical physics that's now been realized.

Back in 1937, Italian physicist Ettore Majorana described a subatomic particle named a Majorana fermion with a unique quantum mechanical state that's resistant to local disturbances. This property made the particle an ideal candidate for building stable qubits for quantum computing, as they'd produce fewer errors.

Microsoft says it managed to observe these particles last year, and has now been able to harness them for practical applications in its chip-building process. Specifically, the company uses Majorana Zero Modes (MZM), which are quasiparticles that exist at the ends of topological superconducting nanowires made from 'topoconductors,' as the building blocks for the qubits in this processor.

The Majorana 1 chip currently fits eight qubits, but can hold more than one million of them
The Majorana 1 chip currently fits eight qubits, but can hold more than one million of them

These 'topoconductors' are a new type of material, which exhibit a whole new state of matter – different from solid, liquid, and gaseous states. In this case, indium arsenide (a semiconductor) and aluminum (a superconductor) are combined, cooled to near absolute zero and tuned with magnetic fields to achieve the state known as topological superconductivity.

This is also where the materials form topological superconducting nanowires with MZM at the wires’ ends. These MZM store quantum information, and share an unpaired electron between them. That makes them invisible to outside interference, protecting the quantum information and allowing for stable operation of a quantum computer.

Naturally, Microsoft isn't done yet. As with other companies building quantum computing tech – like Google – the firm is keen to pack a million qubits into a single processor so it can actually take on complex challenges. That may likely be just years away from now, not decades as we've previously expected.

Source: Microsoft

4 comments
4 comments
Tristan P
Freaky! It's so hard to comprehend how a device such as this could be constructed.
rgbatduke
"cooled to near absolute zero" says it all. It is difficult -- and expensive -- to cool things to the usual superconducting range near absolute zero. This isn't liquid nitrogen, this is liquid hydrogen or helium type ranges. It is difficult and expensive to hold the temperature there. And this -- more than "topology" -- is what makes the extended quantum states stable, as the things that introduce incoherence into a prepared coherent ("entangled", although I do hate that term) multiparticle state are for the most part describable as "thermal fluctuations" -- entropy associated with internal energies well above the ground state of the system.
This is why I remain skeptical of quantum computing. It's right back to things like the Cray, huge liquid-cooled machines maintained by a staff of full-time acolytes, impossible to use without serious ongoing refrigeration and a lot of tweaking. It certainly isn't headed towards my laptop or ordinary cluster computers. This is also the reason that e.g. Penrose is simply barmy when he asserts that human consciousness is due to quantum entanglements at the level of neurons, neurons that are operating at incredibly slow electrobiochemical speeds and that are enormous relative to the speed of light times the probably coherence time of pretty much any extended quantum state at the balmy -- not barmy -- 310 K temperature of the human brain. Atomic and molecular decay times and phonon broadening times at this temperature are order of 1-10 NANOSECONDS, time enough for light to propagate around a meter. Neural impulse transmission, OTOH is centimeters to meters per second. Any quantum information is lost a million to a billion times over in the meantime. Consciousness is a purely macroscale bioelectrochemical phenomenon, quantum only in the sense that all chemistry is ultimately quantum mechanical in nature.
Brian M
Great article and video @rgbatduke Think the cooling close to absolute zero is not really the issue, these are not devices that you need or want (ok maybe!) in your laptop, they are in the environment you would find supercomputer.
There are a lot of smart people and money betting on quantum computing, most of it currently on noisy intermediate-scale quantum (NISQ) computers, that have as much use as a chocolate tea pot, but now we are starting to see companies looking beyond this to large scale chips, the recent Australian investment and now Microsoft.
Things are starting to happen, and it could be an exciting next decade.
As for human consciousness and quantum entanglement not convinced about that either (but never say never), a more reasonable explanation is emergent intelligence from the complexity of the brain and consciousness is really just another word for intelligence!
Ancliff
@rgbatduke Tegmark said exactly that, very politely and technically: Tegmark, M.: 1999, “The importance of quantum decoherence in brain processes”, available at http://arxiv.org/abs/quant-ph/9907009