Sharpening quantum’s predictive powers in the cloud

Interest in quantum computing is on the rise as, step by step, science fiction becomes science fact. Early-stage commercial quantum computers – featuring around a hundred or so qubits – can now be accessed via the web through Quantum-Computing-as-a-Service (QCaaS) providers such as QC Ware, Azure Quantum, Amazon Bracket, and others. QCaaS takes quantum computing to a new audience outside of the research lab, offering users a much more accessible way to explore potentially exponential processing prospects.

Quantum computers encode and manipulate information using a choice of atoms, electrons, or photons. And because these processing bits (or qubits) are quantum in nature, they can represent not just the familiar ones and zeros of classical computers, but also a wide range of possible states in between. What’s more, thanks to quantum entanglement – a phenomenon recognized by the 2022 Nobel Prize in Physics – qubits can perform their calculations in concert and begin to describe complex systems beyond the capability of classical computers.

Qubits in the cloud

QCaaS opens the door to a range of hardware platforms, including quantum annealers (such as those developed by D-Wave) and various gate-based processors. This latter group includes superconducting electric circuits held at cryogenic temperatures – for example, processors designed by rigetti and OQC – as well as trapped-ion quantum computers, such as IONQ’s that manipulate particles with a laser to carry out gate operations. Looking at other architectures, Xanadu’s modular photonic quantum processors use squeezed light pulses to engineer qubits. And there’s no shortage of ingenuity when it comes to designing quantum computers.

It’s an exciting time for stakeholders, but numerous challenges remain due to the sensitivity of the measurements involved. Operators have to go to great lengths to prevent the surrounding environment from spoiling the quantum output. “Qubits are fragile,” Sabrina Maniscalco, CEO and co-founder of Algorithmiq, told TechHQ. “They are imperfect and subject to errors.” Maniscalco leads a team of quantum computing experts based in Finland that is developing ways to help clean up this noise, so that users can make better use of today’s early quantum computers.

Processors inside classical computers are subject to errors too, but over the years, error correction schemes have been perfected that mitigate those missteps. One approach in the quantum computing domain is to build so-called logical qubits comprising a number of physical qubits to engineer fault tolerance. But as images of today’s quantum computers reveal, the hardware is complex and scaling up the number of available qubits isn’t straightforward. With only a relatively small number of qubits currently at their disposal, developers have little capacity for implementing error correction in this classical way.

Quantum capabilities on the rise

No doubt, progress in quantum computing hardware will continue (IBM has a roadmap to ramp up from 127 to 4158 qubits by 2025) and drive progress in qubit-level error correction. But platform-agnostic algorithm processing offers a tool that QCaaS customers can use now to clean up the signal regardless of the type of quantum computer running in the background. Growing interest in quantum-resilient data encryption schemes (driven by the likelihood that powerful quantum computers will one-day break the classical algorithms that protect today’s data as it travels over the internet) points to rising confidence in the capabilities of quantum computers.

“We are in a moment when we can start with quantum and have useful products,” comments Maniscalco. “What makes the difference, is the way of reading out the data so that it’s informationally complete – allowing many properties to be extracted at once.” Maniscalco, like many in the field, believes that quantum computers can have real-world impact across a range of industries. Application areas include materials design and the prediction of molecular structures. With quantum constituents at their heart, quantum computers have the capacity to be much more capable at simulating chemical systems compared with classical machines. And Maniscalco is particularly keen to apply the exponential power of quantum processors to pharmaceutical drug discovery – an area that is ripe for acceleration.

First principles advantage

Writing on the topic of quantum network medicine, her research team highlights that the mechanical simulation of the binding between a small molecule and the biologically active site of a protein target is ultimately a quantum problem. And here quantum algorithms have the potential to describe chemical binding from first principles. This is a powerful predictive feature when you consider that alternative artificial intelligence (AI) led approaches demand huge datasets to map out just a small portion of the measurement space.

The quantum nature of the processing architecture has promise in capturing quantum effects that would otherwise be difficult to model. Examples of such characteristics include molecules that are in a stretched configuration, out of equilibrium in the binding process, which could lead to more realistic and accurate drug models. Big pharma is taking note of the gains being made in the quantum computing sector, both in terms of data processing and usability. And many commentators are describing quantum computing as ‘the next big disrupter’ in the pharmaceutical industry.

Today, there are many more guides available on how to run code on a quantum computer. API’s and SDK’s written in popular computing languages such as Python are also easier to find. Plus, there are startups such Algorithmiq, and others, with quantum computing experts at the helm to provide practical tools and assistance. And as the QCaaS ecosystem grows, users will increasingly be able to focus on the results rather than getting bogged down in the technical details, which can be somewhat mysterious and hard to fathom unless you happen to have a PhD in quantum computing.