Over the past couple of years, there has been a significant interest in Quantum Computing. It is one of the fast-emerging disruptive technology that can significantly impact many industries and societies. BCG estimates that quantum computing will touch $2-$5 billion by 2024, reaching $50 billion by the end of the decade.
The technology is already witnessing exponential growth in its computing power as measured in “Quantum Volume” and is expected to grow further. The “quantum advantage” is likely to be realized in the next few years where real-world problems of consequence can be solved efficiently using a quantum computer that even the most giant supercomputer will not be able to achieve.
Potential applications span multiple sectors including finance, healthcare, manufacturing to supply-chain and logistics. For instance, in banking and finance quantum computing can be used to perform portfolio optimization, risk analysis and fraud detection more efficiently.
Quantum computing can also potentially simplify drug trials by simulating different drug compositions or perform more efficient diagnostics in life science and healthcare domain. In manufacturing, it can be used in designing batteries that can store more energy and in supply-chain and logistics it can be used to track global distributions and do more efficient routing, or predict systemic risks better.
Quantum Computers won’t replace classical computers
To begin with, fifty years of past learning from classical computing has to be incorporated to quicken the integration of quantum and classical systems. Quantum computing is not going to replace classical computing. Quantum computing is good with simulating other quantum systems, finding the lowest eigenvalue of a matrix or doing Monte-Carlo simulations more efficiently.
In short, an application with a specific structure that is computationally intensive is likely to benefit from a quantum computer. This essentially means we will have a “hybrid” model where we will run some “kernel” of an application in quantum hardware while the application itself will run on classical hardware. This is much like using GPUs in the cloud for vector processing. The challenge is to identify those “kernels” that are likely to benefit from quantum hardware.
Quantum computing is a technology born for the cloud; hence it is critical to mobilize developers to code worldwide for Quantum Computing through open-source communities. A big part of the software strategy is to create open-source tools that can be converted to first-class cloud-native components.
It is essential to scale and extend the quantum software to a larger community to take advantage of the architecture while running quantum programs securely and reliably. Users have begun to install and use some of the components from the software stack directly on the cloud architectures. In addition to the cloud-native deployment of quantum workloads, it is critical to break down any potential barriers and democratize access to this new technology.
From drug discovery, health care to finance and logistics
Simulations. Simulating a quantum system is one key value proposition of quantum computing and is particularly useful in the area of drug discovery. The pandemic has brought immense attention to the challenges of discovering a new drug. In general, a drug takes 10-15 years to discover and costs multi-billion dollars to go from development to a product. Simulating a complex molecule is not possible in a classical computer. The number of states required to keep track of all the molecules is humongous; furthermore, there can be many different combinations too in the drug discovery.
Recent years has seen substantial development in both quantum hardware and algorithms. These advances have brought quantum computing closer to impending commercial utility. Drug discovery is one of the promising domains where these applications can enable faster and more accurate characterizations of molecular systems.
Machine Learning. Quantum machine learning has a wide area of potential applications like finance, health science and manufacturing. Diagnostics in healthcare has a significant impact on patient outcome. It is becoming increasingly complex and expensive and some studies indicate error can be up to 25%. Mis-diagnostics can be quite costly not to mention deadly; for example, a study showed that early detection of colon cancer can improve survivability by 9X with 4X reduction in cost.
Classification techniques are used in diagnostics like identifying if cells are cancerous or not, and quantum computing can potentially improve the quality of the results; other applications include medical image diagnostics too. In prediction, quantum computing can significantly improve structural predictions of complex structures like protein, DNA, RNA, etc. Algorithmic developments in quantum machine learning offer interesting alternatives to classical machine learning techniques, which may be helpful for the biochemical analyses involved in the early phases of drug discovery. Other applications include fraud detection in financial transactions, health insurance claims, risk prediction etc.
Optimization. As we saw recently during the pandemic, the supply and availability of oxygen became a huge issue. The technical challenge was how to get oxygen from where it is produced to where it is needed efficiently. These problems in general are computationally intractable, which means that they cannot be solved efficiently in a classical computer. Quantum optimization has the potential to explore the solution space more efficiently and is likely to produce better outcomes.
This is an example of a broader area of supply chain and logistics where quantum computing is expected to make a big impact. Other potential application areas include portfolio optimization in financial services, graph problems like Max-Cut, network optimization in airlines, production yield improvement in semiconductor fabrication to name a few.
In summary, quantum computing can drastically change the computing landscape in the coming decade, much like what GPUs did for AI in the previous decade. Quantum hardware is evolving very rapidly, and quantum software is maturing too. Many use-cases are being explored across many verticals like healthcare, life sciences, finance, supply chain/logistics, etc. These explorations are advancing just as rapidly and the goal is to be ready to take advantage of hardware as and when it is sufficiently powerful. These are exciting times, not only for the technology, but also for computing in general. We are on the cusp of a revolution, and it will be pretty interesting to see how this unfolds in the coming years.
About the authors: Gargi Dasgupta is the Director, IBM Research India and CTO, IBM India and South Asia // Shesha Shayee Raghunathan is a Senior Engineer and IBM Quantum Distinguished Ambassador, IBM Systems