Application-Specific Quantum Hardware is the Most Promising Approach for Early Practical…

by Fabio Sanches, Chiara Pelletti, and Alexei Marchenkov

Quantum computing will dramatically increase our computational power and is projected to generate significant value across many industries — not because quantum logic gates are much faster than logic gates in modern digital processors, but because they are entirely different operations built on different principles. Quantum computers can be more efficient for specific, highly complex practical problems, but they are not a solution to speed up generic computations.

Quantum computing processors have already hit the number of quantum bits (qubits) for which classical simulation of their operation is no longer possible. These new developments in quantum computing technology indicate we have entered the era of Noisy Intermediate-Scale Quantum (NISQ) computers. These are current and near-term systems with up to a few thousand imperfect qubits with noise-limited levels of control and coherence, which restricts the number of gates we can sequentially run in a quantum processor.

For practical applications, NISQ computers must be used as accelerators working in conjunction with high-performance computing systems. Currently, quantum algorithm developers are primarily focused on efficient utilization of such hybrid quantum/classical systems. Precise resource estimation for each algorithm is highly dependent on the specific hardware platform running that algorithm, including architecture and the quantum gates available on that platform. The few general-purpose NISQ computers and compilers available through the cloud, provide a very limited choice of processor architectures and gates. For companies trying to make business decisions on their quantum computing strategy, the hardware architecture specifics are important.

Since 2018, Bleximo has been developing superconducting application-specific systems. To do this, we co-design software and hardware, including processor architecture, gates, compilation, and control methods. The result is a system optimized for specific high-value applications. This strategy also requires us to plan how quantum processors are integrated with their control systems and how to efficiently control the workflow between various parts of the system depending on their intended use. Our perspective is that application-specific systems utilize NISQ technologies in the most efficient way and is the most natural approach for early practical quantum computing applications ahead of the eventual arrival of error-corrected quantum computers.

Originally published at https://medium.com on February 9, 2022.