Scaling up to Problem Sizes: An Environmental Life Cycle Assessment of Quantum Computing

TL;DR

This paper assesses the environmental impact of quantum computers using Life Cycle Assessment, comparing them to supercomputers, and explores how scaling and error correction influence their sustainability.

Contribution

It establishes an environmental profile for superconducting qubit quantum computers and compares it to supercomputers, highlighting scaling effects and error correction impacts.

Findings

Quantum error correction hardware has high environmental impact due to electronic components.

Quantum computers can have environmental advantages at larger problem sizes under certain conditions.

Scaling and error correction influence the environmental performance of quantum computing.

Abstract

With the demonstrated ability to perform calculations in seconds that would take classical supercomputers thousands of years, quantum computers namely hold the promise of radically advancing sustainable IT. However, quantum computers face challenges due to the inherent noise in physical qubits, necessitating error correction for reliable operation in solving industrial-scale problems, which will require more computation time, energy, and electronic components than initial laboratory-scale experiments. Yet, while researchers have modeled and analyzed the environmental impacts of classical computers using Life Cycle Assessment (LCA), the environmental performance of quantum computing remains unknown to date. This study contributes to filling this critical gap in two ways: (1) by establishing an environmental profile for quantum computers based on superconducting qubits; and (2) by comparing it to a functionally equivalent profile of a state-of-the-art supercomputer. With the comparison based on the problem size, the paper shows how the usage time can drive an environmental advantage for quantum computers under specific scaling conditions and quantum error correcting codes. The results emphasize that quantum error correction hardware has a substantial environmental impact due to the numerous electronic components needed to achieve 100 logical qubits. This paper can serve as a basis for designing more environmentally friendly quantum computers and for establishing their environmental profiles, as well as those of the human activities that will use them.