Quantum computing firm QuEra has announced ambitious plans to deliver a 'fault-tolerant' quantum computer by 2028, a development that could mark a significant turning point for the nascent technology. The company’s proposed machine, named Libra, is designed to overcome the critical hurdle of error-proneness that currently limits the practical applications of quantum computers in areas such as chemistry, materials science, and drug discovery.
Currently, a fully functional fault-tolerant quantum computer, capable of detecting and correcting its own errors, does not exist. Yuval Boger, at QuEra, likened the achievement to 'breaking the sound barrier' in computing. Libra is projected to contain between 10,000 and 15,000 qubits, the basic components of quantum computers. These will be organised into 256 'logical qubits', each designed to commit an error only once in a million operations, even if the individual qubits within them are less reliable. This design aims to enable computations of unprecedented complexity.
QuEra plans to make Libra accessible to users through cloud services, in collaboration with Amazon Web Services (AWS). Researchers at the firm anticipate that this will allow Libra to perform a 'megaquop', or one million operations. Quantum computing expert John Preskill from the California Institute of Technology previously suggested that a megaquop machine could herald a new era for quantum computing, expanding its utility far beyond current capabilities.
The journey to deliver Libra by 2028 involves substantial engineering challenges. While the largest array of neutral atom qubits currently stands at 6,100, these have yet to be used for computations. Furthermore, the maximum number of error-corrected logical qubits created so far is 48. QuEra states it is currently running five experimental machines to refine various aspects of Libra, including managing the power of numerous lasers and replacing faulty atoms.
Jonathan King from Atom Computing, another developer of neutral atom quantum computers, highlighted that moving from laboratory demonstrations to a fully functional system requires the integration of numerous scientific and engineering advances. While industry giants like IBM anticipate offering fault-tolerant quantum computers around 2029, QuEra's timeline is notably aggressive. Thomas Wong of Creighton University acknowledges the plausibility of QuEra's 2028 target but also suggests the possibility of delays, while Joe Fitzsimons from Horizon Quantum Computing notes QuEra’s strong track record in error correction breakthroughs.
Should QuEra succeed, a megaquop machine like Libra could be instrumental in complex simulations across physics and materials science, areas where current conventional and quantum computers struggle. Boger also hopes it will accelerate the development of advanced quantum computing algorithms for future generations of fault-tolerant machines, potentially leading to breakthroughs that are currently unimaginable.