Solving the unsolvable: A quantum boost for supercomputing

Quantum computing alone may not be a panacea to all computing problems. However, when paired with high-performance computing, it may be able to find better answers to some of our most pressing issues.

Quantum computing and high-performance computing (HPC) may seem like competitive approaches to solving complex problems, as they’re both designed to perform calculations at unmatched speeds, but they are in fact naturally complementary technologies. Each has different strengths, and using them together can provide much faster answers to problems that were once too big and complex to solve than either can alone.

So what does a world where high performance computers (HPCs) and quantum computers work together look like? The reality is that classical computers are very good at certain tasks that will never be well suited for a quantum computer. But used together, there is tremendous potential for the two computing paradigms to complement and enhance each other’s capabilities.

Quantum vs. HPC

A quantum computer harnesses a number of physical phenomena not available to classical computers. While classical computers only allow bits of information to live in one state (0 or 1) at a time, a quantum computer uses qubits (or quantum bits) that enable bits of information to be a 1, 0 or both 0 and 1 simultaneously. As such, a quantum computer is able to compute multiple possibilities simultaneously in order to find the optimal solution. This leads to fundamentally different uses than those best suited to HPC systems.

A quantum computer excels at tasks that require it to determine optimal solutions among a broad range of possibilities, problems like design computation, Monte Carlo simulation and discrete optimization. By exploiting quantum parallelism, a quantum computer is able to evaluate a much broader range of possible in much less time. Because a quantum computer is probabilistic rather than deterministic, the computer returns many very good answers, providing not only a single answer, but also other alternatives to choose from. In this way it can help narrow the search space for problems with huge numbers of possible answers like finding the needle in a very big haystack.

HPC systems are very well suited for well-defined, data-crunching numerical simulations, for example, weather forecasting, nuclear weapons modeling, and molecular simulations. Using the range of answers returned by the quantum system, the HPC system can evaluate each of the possible solutions very quickly to come up the best answer to the particular problem at hand. Therefore, the inherent nature of classical systems is actually beneficial here.

Quantum and HPC

As it turns out, many of the world’s most complex problems require a combination of simulation computations (which are better suited for classical computers) and discrete optimization problems (which are better suited for quantum computers.)

There are several areas of research where a quantum computer can be used along with HPC to improve outcomes in multiple disciplines. This ranges from formulating radiotherapy treatment plans for cancer patients that minimize collateral damage, to predicting the movement of oil in a reservoir and determining how to extract the oil most efficiently. These real world problems are just two examples of how the combination of simulation computation by HPC and optimization computation by quantum computers can be applied to real-world problems.

Quantum computing technology is poised to become a powerful option for certain complex problems. High performance computing already has a solid infrastructure in certain areas, and it’s had the time to develop immense strength. Together, these two technologies can expand their reach to new ground, and find better answers to some of the most pressing problems of our day.

Gigaom