Your questions – answered.
Everyday computers speak in a language that we call binary. No matter what you do on a computer, whether it’s checking email or building complex computational models, it all is encoded in a sequence of these binary operations. Some operations, however, are difficult to carry out using this binary language. The advantage of quantum computers is that they speak in a more complex and rich language that classical computers are not good at speaking. Many aspects of linear algebra are native to this quantum language, which makes them better than classical computers at some of these problems.
Today we can manipulate billions of classical binary digits (“bits”) very quickly, but only dozens of quantum bits (“qubits”). Even with not-so-many qubits, quantum computers will have an advantage over classical computers in certain classes of algorithms. However, for problems that are easy for classical computers in the first place—say, adding two numbers together—there is no advantage in using quantum computers. This means that quantum computers function as a co-processor, like a GPU. The classical computer does most of the book-keeping, and only certain extra hard jobs are tackled by quantum computer.
The small quantum hardware devices today are referred to as noisy intermediate scale quantum (NISQ) computers. To date, NISQ computers have not done much in the way of useful problem solving, but the improvement of these machines is already approaching the point where they will compete with the best that classical supercomputers can offer.
Today’s fully quantum computers and quantum simulators are large enough to explore potential use cases by integrating them with classical computers. While hybrid solutions might demonstrate quantum advantage, most applications will more likely require additional hardware development before this is possible. It is nonetheless possible to benchmark these solutions on smaller versions of the problem to determine whether the advantage is expected.
Quantum-inspired solutions, on the other hand, make it possible to solve larger instances. These solutions, though lacking the full power of quantum computing, approach problems in a similar way and may offer advantages over legacy methods. In addition, their design offers a straightforward translation to fully quantum computers and makes them a useful window into their performance.
Quantum computers are known to be efficient for factoring numbers. They can efficiently simulate solid-state compounds and molecules, enabling the design of new chemicals, materials and drugs. Quantum scientists have also built heuristics for solving hard instances of optimization problems. See many more use cases here: [Link to our list of use cases] Many of these solutions still require significant improvements to quantum hardware technology before they will be available.
Most experts predict that within the next few years we will have a large enough quantum computer to enable an advantage for some real business problems. However, it is still unclear exactly which problem it will be. Most expect that quantum advantage will first be achieved in one of three areas: machine learning, simulation of quantum materials (e.g. chemistry), or optimization.
Our team put together these posts to help people understand what we do.