A fault-tolerant quantum supercomputer is a step closer according to Microsoft. A new roadmap from the Microsoft Azure Quantum team sets out the six steps required to achieve the goal of making a machine capable of solving some of the world’s most complex problems. Microsoft also unveiled a new AI “Copilot for quantum” and Quantum Elements, a product that merges HPC, AI and quantum to speed up scientific discovery.
The roadmap builds on the first creation of topological qubits last year. These are seen as a requirement for true, fault-tolerant quantum computers as they are less prone to the noise and errors of other forms of qubits. Microsoft says it uses particles known as Majorana in a topological state to form the basis of its qubits. “It’s akin to inventing steel, leading to the launch of the industrial revolution,” claims Krysta Svore, Microsoft’s vice president of advanced quantum development.
Microsoft says logical qubits, formed from many physical qubits, are required for a true quantum supercomputer. The more stable the qubit you start with, the easier it is to scale up towards supercomputer levels as you need fewer physical qubits per logical qubit. The company says it has experimented with all forms of qubits including spin, transmon, gatemon and others but none scale effectively.
Its engineers found that to have a truly stable and scalable quantum supercomputer “topological qubits” were required. To achieve this a breakthrough was needed in physics that had eluded scientists for nearly a century. Microsoft achieved this last year with the creation and manipulation of matter in a topological state. In this state, qubits are more easily manipulated, more stable and have a smaller footprint allowing for greater scale.
The next stage for Microsoft on the road to a quantum supercomputer is to create “hardware-protected qubits”, otherwise known as a topological qubit that has built-in error protection and can scale to support a reliable qubit. Each of these qubits has to be smaller than 10 microns to fit a million on a credit-card-sized chip and be fully controllable.
The other stages include improving the quality of those hardware-protected qubits to allow for entanglement and reduced error rates. Then the number of qubits needs to be scaled up and put together into a programmable QPU. The final stages are work on resilience and building towards what Microsoft calls rQOPS, a new metric to track reliable Quantum Operations Per Second.
There are no timeframes on when this might be achieved, but Microsoft Azure says it expects the entire roadmap to be completed in “years not decades”, so we may see its launch before the end of this decade.
Microsoft Quantum Elements: three stages of quantum development
Microsoft splits the quantum era into three stages, starting with the current NISQ, or noisy stage of quantum computers that are error-prone and small-scale. “The ultimate unlock will come when organisations are able to accurately design new chemicals and materials with a quantum supercomputer,” the company says. To get there, the industry will follow a similar path to that of the development of classical supercomputers, moving from vacuum tubes to transistors and on to integrated circuits placed at scale.
The next stage is resilience, as quantum systems begin to operate on reliable logical qubits that are stable enough to perform true calculations. As this is established, the final level will begin as quantum supercomputers scale to a level and potential not possible with current or future classical supercomputers.
Satya Nadella, Microsoft CEO, said the goal is to “compress the next 250 years of chemistry and material science progress into the next 25”, hinting at the scale of compute power possible with a quantum supercomputer. Part of this will come in through the announcement of Azure Quantum Elements, a combination of quantum, AI and high-performance computing to speed up scientific research.
This, according to Azure, allows developers and scientists to reduce R&D time and prepare for scaled quantum computing. It claims some customers are seeing a speed up of certain chemistry simulations by 500,000 times – compressing a year of compute into a minute.
AI and quantum are entwined
“Chemistry is in everything,” said Ansgar Schaefer, a vice president at BASF, one of the first users of Quantum Elements. Schaefer leads the company’s quantum chemistry research, and said: “To be able to improve products and processes, it’s really about understanding the chemistry behind them on a microscopic level. And the more complex the challenge, the more computing power is required. [Azure Quantum Elements] is a tool that gives us the additional required capacity to help advance completely new research approaches and increase the efficiency and speed of development.”
Quantum Elements incorporates Microsoft AI chemistry models and runs them in a hybrid quantum/HPC environment. These can also be instructed using Copilot, Microsoft’s foundation model natural language AI tool.
Copilot has already been launched throughout the Microsoft product family including Windows, GitHub and Microsoft 365. “Copilot in Azure Quantum helps scientists use natural language to reason through complex chemistry and materials science problems,” the company explained.