Quick, here comes Quantum

The evolution of quantum computing, which is expected to result in an exponential increase in speed and complexity of problem-solving, has reached a watershed.

“We are at a very pivotal moment in quantum computing,” said Krysta Svore, who leads the quantum computing team at Microsoft Research. She was speaking to Business Times after the unit announced a breakthrough described as “a crucial milestone on our path to building a hybrid supercomputing system that can transform research and innovation across many industries”.

Quantum computing promises to solve problems in chemistry and material science that are currently beyond the reach of digital technologies.

“Our mission is to accelerate scientific discovery,” said Svore. “We want to take the next 250 years of scientific discovery and compress that into the next two decades because we really need to help save and feed our planet.”

According to Svore, this ambitious goal relies on harnessing cutting-edge techniques across AI, high-performance computing (HPC), and quantum computing to transform numerous scientific fields.

To achieve this, Microsoft has developed a platform called Elements, which Svore described as a “computational workbench.” This platform combines AI, HPC, and quantum computing capabilities within the Azure cloud environment, enabling rapid scientific discoveries that would otherwise take decades.

Svore provided a striking example of this platform in action through Microsoft’s collaboration with Pacific Northwest National Laboratories.

“We were able to discover a new battery electrolyte by using AI and HPC pieces of Elements. We generated 32-million candidate battery materials, filtered them down via AI and advanced HPC-based simulations to 20 candidates, and then synthesised one in the lab. This process took just 80 hours, which would have taken roughly 20 years in the past.”

The new material not only worked as a battery but also used 70% less lithium, replacing it with the more abundant and less expensive sodium.

Svore emphasised that such advancements were just the beginning, as quantum computing keeps evolving. It promises to solve problems in chemistry and material science that are currently beyond the reach of classical digital technologies.

Where traditional computers use bits as their basic units of information, which can only be represented by a 0 or 1, quantum computers use “qubits”, which can exist in multiple states simultaneously, enabling vastly more powerful and efficient data processing.

“Quantum computing is about delivering better accuracy on understanding chemical reactions. We need to understand everything about the electrons – what they’re doing, how they’re behaving, and how they’re interacting. A lot of that requires quantum mechanics.”

However, achieving these breakthroughs requires scaling quantum computing to a level where it can handle these complex problems.  Currently, quantum computers have achieved a little more than 1,000 qubits.

“Ultimately, we need to build a quantum machine with upwards of a million physical qubits. Recently, we had several breakthroughs, including the creation of the most reliable logical qubits on record.”

Svore detailed the three stages of quantum computing development. The first stage, called Noisy Intermediate Scale Quantum (NISQ), involves creating and operating physical qubits. The second stage, resilient quantum computing, uses several physical qubits to form a unit called a logical qubit, which is more stable and error resistant. The third and final stage involves scaling up to thousands of logical qubits, which requires millions of physical qubits.

“In April, we announced that we had graduated from level one to level two. Together with (quantum software developer) Quantinuum, we demonstrated four logical qubits that have an 800 times better error rate than the corresponding physical qubits.”

Svore said that the integration of quantum computing with AI and HPC is essential for addressing complex scientific problems. “We want to bring quantum computing in to help accelerate scientific discovery. This is critical for solving problems like improving industrial catalysts, making the processes more efficient and sustainable, and ultimately helping to address global issues such as food scarcity.

“In artificial fertiliser production, you’re relying on a process called nitrogen fixation, which happens in the soil all the time, very naturally. It’s happening in little microbes in the soil. We want to be able to mimic what the microbes are doing. If we can model that computationally, and to do that with high accuracy, then we can design an industrial catalyst to mimic what those microbes do in the soil. Then ultimately, you can look at enabling the creation of artificial fertiliser in more remote areas of the world, where you’re not having to transport it. This can then help address issues of food scarcity, and improve the whole fertiliser production process.

“From food scarcity to sustainable energy sources, these are the types of problems we can address with a quantum machine at scale.”

The logical qubit breakthrough has also allowed Microsoft to position itself more aggressively against other major players in the quantum world. The world is seeing significant investments and global interest in quantum computing, such as the Australian government controversially announcing it will provide A$940 for the US-based quantum-startup PsiQuantum to build a fault tolerant quantum computer (FTQC) in Brisbane. The investment, from the country’s National Quantum Strategy budget, makes PsiQuantum the world’s most funded independent quantum company.

Svore said this was in line with Microsoft’s own ambitions: “We are well on our way, with level two capabilities already here. We are fully committed to building a quantum machine at scale. We’ve identified that we need a unique qubit approach – one that considers size, speed, and controllability. That’s why we’ve invested in topological qubits,” she said, referring to qubits that can encode information and are practical for scaling up quantum computing. They are also a path to reaching level three quantum computing.

“We believe it’s a matter of years, not decades, to reach level three. Graduating from level one to level two is a significant step, and we are on the path to level three.”

Microsoft will shortly make its logical qubits available for private preview through a system called Azure Quantum Elements, designed to accelerate scientific discovery. The targeted customers include those in chemistry, material science, and pharmaceuticals.