Quantum Utility, Advantage and Supremacy

- Overview

Quantum computing, an innovation most can’t define and still don’t properly understand, might be the next obscure technology to have a seismic effect on business. Quantum computing applies the laws of quantum mechanics to simulate and solve complex problems that are too difficult for the current genre of classical computers.

Quantum utility, advantage, and supremacy are three classifications of how useful a quantum computer can be:

• Quantum supremacy: A quantum computer can solve a task faster than any classical computer, even if the task is not useful. This term originated from the misconception that quantum computing would replace all other types of computing. Examples of problems that quantum computers can solve with supremacy include complex derivatives pricing, large portfolio optimization, and large trade execution. However, quantum supremacy can be broken as classical computers improve.
• Quantum utility: A quantum computer works, but there's no particular advantage to using it over a classical computer. This concept demonstrates the effectiveness and practicality of quantum computers for various applications.

The current field of quantum computers isn’t quite ready for prime time: McKinsey has estimated that 5,000 quantum computers will be operational by 2030 but that the hardware and software necessary for handling the most complex problems won’t be available until 2035 or later. Yet organizations need to start thinking now about where they might leverage the technology to solve real-world business problems. 

- Quantum Utility vs. Quantum Supremacy

Quantum utility and quantum supremacy are two ways to measure how useful a quantum computer can be.

Quantum utility is the ability of quantum computers to solve problems that are beyond the reach of classical computing methods. Quantum computers can provide reliable and accurate solutions to problems that were previously only accessible through classical approximation methods. 

Quantum utility is important because it allows researchers to: Discover new scientific insights, Verify classical approximations, Explore meaningful scientific problems, and Benchmark classical methods.

A long-sought goal is to achieve quantum "supremacy" - proving that quantum computers can solve calculations that conventional computers on Earth cannot - but there are no practical benefits. Google claimed to have achieved this through a landmark scientific paper published in 2019, but IBM apparently expressed doubts. Anyway, this is an exercise in computer science and has no practical meaning in the real world.

Since Google's announcement, the industry has stepped up its efforts to achieve quantum "advantage," defined as a commercial or scientific advantage achieved by exceeding the computing power of the largest supercomputers in a relevant application. It's certainly more useful than quantum supremacy as a reference point for comparison and benchmarking. 

Quantum supremacy is often associated with major breakthroughs in drug discovery, financial transactions, or battery development.

- Quantum Supremacy vs. Quantum Advantage

Quantum supremacy and quantum advantage are terms that are often used interchangeably, but they have some differences in definition:

• Quantum supremacy: When a quantum computer solves a problem that is impossible for a supercomputer to solve in a reasonable amount of time. This term was coined in 2012 by John Preskill, a theoretical physics professor at Caltech. Achieving quantum supremacy in a practical way would require a quantum computer with millions of qubits, but the largest machine currently available only has about 1,000 qubits. 
• Quantum advantage refers to quantum computers solving tasks that have practical usage, for example, problems stemming from physics or economics.
• Quantum computers can perform complex calculations at unprecedented speeds by processing information in parallel using quantum algorithms. Quantum advantage could have potential applications in machine learning, optimization problems, and complex simulations. 
• “Supremacy” does not require quantum error correction (QEC), whereas “advantage” requires QEC for the results to be useful. “Supremacy” has no relevance for commercial activity, whereas “advantage” is highly relevant for commercial activity. 
• “Supremacy” can be demonstrated on any problem, regardless of its usefulness, whereas “advantage” is concerned with useful, real-world problems. “Supremacy” does not require quantum error correction (QEC), whereas “advantage” requires QEC for the results to be useful.

- The Ultimate Goals of Quantum Computers

As the quantum computing industry continues to move forward, so does the goal.

• Quantum Supremacy: is the ability of a quantum computer to solve a task faster than any classical computer. The stress here is on 'a task', because it can be anything. The definition also doesn't include any word about the usefulness of the task performed. Additionally, as classical computers get better, quantum supremacy can be broken.
• Quantum Advantage: refers to the point where quantum computers start to solve practical problems faster or more efficiently than classical computers. This is not just about theoretical superiority but about providing tangible benefits in real-world applications.
• Quantum Utility: is the stage where quantum computers are used for practical, economically viable applications, providing clear and tangible benefits over classical computing methods.

Getting quantum computers into use as soon as possible is a much better goal than quantum superiority and quantum advantage. The ability of a quantum computer to far exceed the capabilities of a classical supercomputer is known as quantum supremacy. A quantum device that can perform a calculation better than a supercomputer at a certain task is called quantum advantage.

However, quantum supremacy misses an important point: should we really wait for a million-qubit quantum steampunk gold chandelier to overtake a supercomputer before thinking quantum computers make sense? Or should we focus on measuring performance improvements compared to hardware units used in today's classical computers, such as individual CPUs (central processing units), GPUs (graphics processing units), and FPGAs (field programmable gate arrays)?

Because for this still nascent industry, the more worthy goal may be to achieve quantum "utility" or usefulness as soon as possible. Quantum utility is defined as a quantum system that outperforms a classical processor of comparable size, weight, and power in a similar environment. 

- Accelerating Quantum Commercialization

Those who have studied quantum computing deeply know that it will have a huge impact on IT, business, economy and society. The future of quantum supercomputing mainframes with exponential acceleration, error-correcting qubits, and a quantum internet will be a very different world than the one we live in today. 

That said, similar to the classic mainframes of the 1960s, quantum mainframes are likely to remain large and fragile machines for the foreseeable future, requiring ultra-low temperatures and complex control systems to operate. Even when fully operational, there will only be a handful of quantum hosts in supercomputing and cloud computing facilities around the world. 

The quantum computing industry would be better off if it mimicked the success of classical computers. When personal computers came out in the late 1970s and early 1980s, IBM and other companies were able to introduce new models each year that offered incremental improvements over previous models. This market dynamic drives the development of Moore's Law. 

Quantum computing needs similar market dynamics to scale and thrive. Investors can't be expected to keep throwing money at them, waiting for quantum computers to overtake a handful of supercomputers. The annual release of new, improved and more "useful" quantum computers will provide the assurance of revenue, driving the long-term investments needed to realize the technology's full potential. 

With a steady stream of useful quantum systems available for a variety of applications, there's no reason to wait in line to handle computations on one of the few large-scale quantum hosts available in the cloud when you can have a quantum processor right next to you. You, integrate with your existing classic systems. Your application may require instant computing that "quantum in the cloud" cannot deliver in time, or you may have to rely on local or on-board computing if cloud access is not available.

<More to come ..>