Microsoft’s Quantum Computing Journey

Microsoft recently made several announcements relating to its quantum computing journey during the recent Accelerated Scientific Discovery virtual event. While most of them were natural corollaries to activities that they had shared from time to time, there were two that instantly turned a few heads in the world of technology. 

The first of these related to Microsoft’s announcement that they had crossed the first milestone in a six-step roadmap towards a quantum supercomputer by publishing a peer-reviewed research paper demonstrating the achievement. Of course, the second related to integrating its AI-powered assistant Copilot into Azure Quantum. 

So, what is quantum supercomputing now?

The tech giant has defined a quantum supercomputer as a quantum system capable of solving problems that most powerful classical ones cannot. While quantum computers come with physical qubits, a quantum supercomputer needs reliable logical qubits, each made up of several physical qubits. 

Says Krysta Svore, engineer and VP of advanced quantum development, “A quantum supercomputer will be one of the most consequential technologies of our generation, enabling us to compute more like nature and to solve many of society’s most difficult challenges.” The first step towards building one involves developing a new qubit with inherent stability at the hardware level itself. 

What’s made the heads turn?

In the past, Microsoft has pointed out that the traditional qubits used in several intermediate scale quantum computers do not have the required scalability built in. This caused the company to engineer a new qubit, which has been validated by a peer-reviewed paper published in “Physical Review B,” a journal of the American Physical Society. 

The paper reveals the development of a device that can induce a topological phase of matter characterized by Majorana Zero Modes or MZMs that paves the way for creating a more stable qubit. Svore says that Microsoft engineered several types of qubits but none were capable of scaling all the way to a quantum supercomputer.  

“That’s why we’re engineering a brand new type of scalable qubit that is inherently more stable than other qubits, and is small, fast and controllable,” Svore says, adding that the company can now create and control MZMs. “It’s akin to inventing steel leading to the launch of the Industrial Revolution. This achievement clears the path to the next milestone, a hardware-protected qubit that can scale, which we’re engineering right now,” she says.

From physical qubits to a qubit network

The official says traditionally the industry had measured quantum systems by counting the physical qubits or quantum volume, but to measure a quantum supercomputer’s performance would require understanding how reliable the system will be for solving real problems. For this,  the machine would need to deliver at least a million reliable quantum operations per second with an error rate of 10 to the power of minus 12 operations. 

Readers would recall that Microsoft had set a six-step process to achieve its dream. Now, in a blog post it suggests that these involve the following: (a) Creating and controlling Majoranas, (b) The development of hardware-protected qubits (c) High-quality hardware-protected qubits (d) A multi-qubit system (e) A resilient quantum system, and (f) A reliable quantum supercomputer. 

The official highlighted during the virtual event that the fifth step would be foundational in nature and a breakthrough that could enable the first reliable quantum operations per second mark. “We will have logical qubits for the first time and a resilient quantum system. Once we have these reliable, logical qubits, we [will be] able to engineer a quantum supercomputer,” she said.

Integrating AI for Copilot in Azure Quantum

Meanwhile, the second announcement integrating AI for Copilot into Azure Quantum indicated that the company was indeed on a warpath. CEO Satya Nadella had set the goals on quantum supercomputing by stating that ““Our goal is to compress the next 250 years of chemistry and materials science progress into the next 25.” 

One part of the strategy involves integration with artificial intelligence or AI. Microsoft has been integrating OpenAI’s generative pretrained transformers into its product line, and this time it’s extending it to Azure Quantum. The Copilot is designed to help scientists use natural language to reason through complex chemistry and materials science problems. 

And Svore is quite enthused by this development. “You can ask Copilot questions about quantum. You can ask it to develop and compile quantum code right in the browser. You can ask about chemistry and molecular properties, and you can quickly get up to date on research. Our aim is to encourage stronger bridges between these transformational fields for learners and innovators alike.”