Making Quantum Computing a Reality

Scientists have theorized about the potential of quantum computing — that is, a new approach to computation that uses probabilities, rather than binary signals, to make calculations — for decades. But in recent years, both private and public sector investment into developing quantum computers has grown significantly, with one report projecting investments of more than $800 million in 2021 alone.

Quantum technology could revolutionize everything from genomic sequencing to transport route optimization, from code-breaking to new materials development. But while quantum computers exist in the lab, general-purpose quantum computers aren’t yet available for commercial use. How can businesses respond to potential disruptions from this technology before it has actually emerged into the mainstream market?

To explore this question, it’s helpful to look to historical examples of major technological transitions, such as the shift from analog to digital photography, or from internal combustion to electric engines. In many of these cases, companies leveraged a hybrid approach to integrate new technologies: Rather than attempting to switch over to the new technology all at once, they developed products that combined elements of old and new technologies. For example, the hybrid-electric Prius enabled Toyota to learn about making electric cars while still leveraging its foundation of expertise with traditional gas engines. After launching this initial hybrid model, Toyota moved forward with plug-in hybrid cars and fuel cell electric cars, paving the way for its eventual launch of all-electric cars several years later.

So, what might a similar hybrid approach look like for quantum computing? One organization that’s been investing substantially into quantum computing is Infosys, and so we reached out to several researchers and business leaders at the company to learn more about their work. Through a series of in-depth interviews, we found that Infosys has been experimenting with two hybrid approaches to begin commercializing existing innovations and build a bridge to the future of quantum computing:

• Quantum-inspired algorithms: algorithms that run on classical computers, but whose logic is based on emulating various quantum phenomena (in contrast to the binary, deterministic approach of traditional computing)
• Hybrid models: models that enable classical computers to work together with quantum computers

Infosys has been leveraging these approaches in many different fields, both independently and in partnership with startups. Below, we describe three key applications of quantum computing in which Infosys has begun investing: optimization problems, in which the company has been exploring the potential of quantum-inspired algorithms, and machine learning and cybersecurity solutions, in which Infosys has begun leveraging hybrid models.

Optimization Problems

While classical algorithms are effective in many domains, they can be prohibitively slow and expensive when it comes to solving certain kinds of optimization problems. For example, in finance, it is difficult to use traditional computers to optimize portfolios, since this necessitates rapid, real-time analysis of the constantly fluctuating risk values associated with investing in each individual stock. To address this challenge, Infosys developed quantum-inspired algorithms to optimize the selection and allocation of assets. This enabled the company to build a diversified portfolio that maximized returns and minimized risks for more than 100 stocks in just one minute, ultimately achieving a 21% improvement in returns compared to conventional (i.e., non-quantum-inspired) asset allocation strategies.

Another area in which traditional computers can struggle to optimize accurately and cost-effectively is in supply chain. To explore the potential for quantum computing in this space, Infosys partnered with QpiAI, a startup developing quantum-inspired solutions for supply chain optimization. While these projects are still in development, the team has already shown that its algorithms enable a 60% cost reduction in vehicle routing optimization.

Machine Learning

Machine learning algorithms depend on highly intensive (and expensive) computation power to extract learnings from large datasets. Especially when it comes to analyzing datasets that are highly imbalanced — that is, where the cases you care about identifying are extremely rare — quantum computing could both dramatically reduce costs and improve these models’ effectiveness.

In financial fraud detection, for example, the number of fraudulent transactions is tiny compared to the number of normal transactions. This makes it hard to develop classical machine learning algorithms that can identify fraud sufficiently quickly and accurately. But Infosys took a hybrid approach, building out a hybrid neural network algorithm in which most network layers used classical computing, while some layers incorporated input from a quantum computer. With this system, Infosys was able to achieve a 1.66% improvement in the accuracy of its fraud detection tool — a difference that may seem small, but has the potential to translate to significant savings given the massive scale of the global financial system.

Cybersecurity

Current cybersecurity protocols typically use pseudo-random numbers to encrypt sensitive information such as passwords, personal data, or even blockchains. The problem is, quantum computers can easily crack the methods traditional computers use to generate random numbers, potentially posing a huge threat to any organization using these standard encryption tools. Yet, alongside this new threat, quantum technology also holds new possibilities: Quantum systems can produce a large, reliable stream of “true random” numbers that cannot be decrypted with either classical or quantum systems.

Infosys partnered with quantum cybersecurity firm Quintessence Labs to develop a hybrid solution that first generates true random keys with a quantum random number generator, and then funnels those keys into classical cryptographic algorithms and encryption systems. This approach makes it possible to generate truly random, unpredictable numbers for use in a wide variety of existing commercial applications, enabling a new level of cybersecurity for any organization that works with large quantities of sensitive data.

.  .  .

These applications might sound like science fiction, but they are very real. While quantum computers still have a long way to go before they’re ready for prime time, businesses are already leveraging quantum technologies in hybrid solutions, blending the old with the new to build a bridge between the reality of today and the potential of tomorrow. Investing in this hybrid strategy now is the best way for companies to develop the expertise in quantum principles and software development that will become critical as these technologies reach maturity. It also means that regardless of how exactly quantum hardware develops and which platforms ultimately emerge as industry standards, the algorithms being developed today will be able to operate on almost any type of quantum hardware (rather than being limited to just one system). Ultimately, taking a hybrid approach enables companies to serve customers today, while getting a leg up on the future — even if some of the technology involved is still catching up.