Quantum Chip Milestone Stokes Excitement for the Future of Computing

A New Quantum Computing Milestone

Quantum computing is, together with AI, a new horizon for the tech industry. This is because, contrary to other non-silicon computing technologies, like graphene, photonics, or even biological organoids, quantum computing performs calculations in a radically new way.

It does so by leveraging quantum effects, with the calculation not being made with 0 & 1 (binary bits), but qubits, where particle data is either 0 AND 1 at once, or 1, or 0. These different methods mean that for some mathematical problems, quantum computers can be absurdly more efficient than normal computers.

And this is exactly what Google just demonstrated: it has unveiled a new chip that takes five minutes to solve a problem that would currently take the world’s fastest supercomputers ten septillion – or 10,000,000,000,000,000,000,000,000 years – to complete.

Besides this remarkable performance, this might also be the first quantum computer chip that has a reasonable chance to be scalable and reliable for useful computations. The results were published in the top scientific publication Nature, under the title “Quantum error correction below the surface code threshold”.

Meet Willow

The Google quantum computing chip in question is called Willow.

This is the last version of a series of increasingly powerful quantum chips developed by Google, called Foxtail, Bristlecone, and Sycamore.

Source: Google

To test Willow’s performance, the random circuit sampling (RCS) benchmark was used to compare it to a normal supercomputer. This technique was created to measure Sycamore’s performances, which evaluates the amount of qubits required to match an equivalent amount using a classical computer.

Source: Google

Scalability Finally Within Reach

One major limitation of quantum computing so far is that qubits are extremely fragile. This is because quantum states tend to be very short-lived, with interferences from the environment destabilizing the conditions required for quantum calculation.

This is why all current quantum computers use ultra-cold conditions and require superconducting materials. This makes it not only more complex and expensive but very difficult to use reliably.

“To get a quantum computer to perform useful calculations, “you need quantum information, and you need to protect it from the environment — and from ourselves, as we do manipulations on it.”

Julian Kelly -Physicist at Google leading the quantum-hardware division

To add to the problem, so far the more qubits in a system, the higher the error, with any instability causing a domino effect. So the more qubits, the more error, up to the point where adding more computing power actually decreases the overall performance.

And maybe more than on its computation ability, this is where Willow is a true breakthrough.

You can also directly see more about the exact performance of Willow in the Google dedicated video presentation:

Error Correction

The fundamental step for Google has been to work on so-called error correction. The idea is that in a given quantum computer, some qubits will be more reliable than others.

So, the outlier/under-performing qubits can be reconfigured to have them perform like the rest of the chip.

Source: Google

Further improvement can be achieved by improving the calibration method so errors are minimized through software improvement on all qubits.

Quantum Error Correction Threshold

This threshold is a very long sought-after goal of quantum computing scientists. This is where the error rate is low enough that the more qubit you add, the less error you get. This was long theorized but only achieved for the first time with Willow.

We tested ever-larger arrays of physical qubits, scaling up from a grid of 3×3 encoded qubits, to a grid of 5×5, to a grid of 7×7 — and each time, using our latest advances in quantum error correction, we were able to cut the error rate in half.

Google Quantum Computing Department

Source: Google

In essence, instead of compounding on each other until collapse, the errors are now rare enough that they can be corrected with additional computing power, increasing the reliability of the whole.

A key factor for Google to achieve this breakthrough was a radical improvement in a specific qubit performance called “qubit coherence time”. This is the time when the qubit can maintain its coherence state, a prerequisite condition to performing quantum calculation.

Willow achieved 100 micro-second coherence state, up from just 20 micro-seconds for 2019 Sycamore.

This is an important data point, as the improvement in coherence state, leading directly to exponentially more reliable quantum computers, could be seen as the current quantum version of Moore’s law, which has driven the constant improvement in “normal” computers.

As the first system below this threshold, this is the most convincing prototype for a scalable logical qubit built to date. It’s a strong sign that useful, very large quantum computers can indeed be built. Willow brings us closer to running practical, commercially relevant algorithms that can’t be replicated on conventional computers.

Google Quantum Computing Department

All of this was possible thanks to Google’s long-term investment in its superconducting quantum chips manufacturing facility.

From Physical To Logical Qubit

For any useful calculation, quantum computing needs to use logical qubits or the actual computation unit produced by the physical chip.

With Willow’s rate of error, each logical qubit is expected to be made of around 1,000 physical qubits. Further improvement in error correction could even bring this down to 200 physical qubits.

Currently, Willow has a capacity of 105 qubits.

Next Steps

With this breakthrough, the perspective of building a stable, reliable, and scalable logical qubit is now within reach. Especially as adding physical qubit now reduces the error rate exponentially, instead of growing it exponentially.

Most likely, the next step will be to build the successor of Willow with even more qubits, while also improving the error correction technology. Together, this could allow each of Google’s next quantum chip designs to be enough to create a full logical qubit.

This will still leave another challenge, which is to network together the logical qubits so that they can share and exchange quantum states.

Only then will we be able to speak of a true quantum computer that can be routinely used and, more importantly, scaled up at will.

We previously discussed a few of the technologies and parallel breakthroughs that could boost quantum computing performance and architecture beyond improving qubits, for example:

Overall, Google’s breakthrough is truly groundbreaking and a radical evolution of quantum computing prospects, with the first-ever realistic target for a scalable commercial computer.

However, even the company’s lead of  Google’s Quantum AI lab, Hartmut Neven, reminds us that this is not happening tomorrow:

“A chip able to perform commercial applications would not appear before the end of the decade. Initially, these applications would be the simulation of systems where quantum effects are important.

For example, it’s relevant when it comes to the design of nuclear fusion reactors to understand the functioning of drugs and pharmaceutical development, it would be relevant for developing better car batteries and another long list of such tasks.”

Applications

Quantum computing’s potential is massive and could revolutionize virtually every scientific field. A few applications stand out as especially impactful:

• Biochemical modeling: from determining the 3D shape of a protein to gene expression, the calculation of complex biological molecules to the atoms could revolutionize biotechnology research.
• Climate modeling: Climate models are extraordinarily complex and stretch the limits of what current supercomputers can do. A better understanding of the climate, with a finer calculation scale in the model, both geographically and in time, could help in understanding climate change risks.
• Semiconductors: Quantum computers could be used to make normal computer chips a lot more powerful. With “normal” chips now reaching the nanometer scale, quantum phenomena become increasingly problematic, and quantum computers might be needed to solve them.
• Material Science: Understanding quantum physics better and the reaction of materials down to individual atoms can open new designs for materials used in aerospace, batteries, 3D printing, manufacturing, etc.
• Cryptography: Quantum computers could potentially make all current cryptography methods obsolete. This is a serious concern for military, financial & IT systems. But at the same time, it could make cryptography even more secure.