Quantum X Labs has detailed a path toward achieving approximately 800 logical qubits and over 100,000 physical qubits by 2030, signaling ambitions that extend beyond simply building a fault-tolerant computer. The company recently benchmarked its quantum error-correction decoder against others using NVIDIA CUDA-Q QEC and Google’s surface-code hardware dataset, achieving the lowest logical error rates in the process.

“A quantum computer is measured by how many reliable logical qubits it can deliver,” says Prof. Nir Sharon, Chief Quantum Scientist of Quantum X Labs, emphasizing that the roadmap focuses on physical qubits, logical qubits and gate performance as key metrics.

Cold-Atom Roadmap Targets 800 Logical Qubits by 2030

Quantum X Labs is targeting a 99.999999% gate performance by 2030 as a key milestone in its roadmap for building a fault-tolerant quantum computer. This ambitious figure, detailed in the company’s recently published plan, highlights a focus on quality of qubits rather than sheer quantity, a strategy the company believes will determine commercial viability. The roadmap outlines a progression from the company’s current 50+ physical qubit cold-atom platform toward error-corrected computing, with intermediate goals set for 2027 and 2029.

2027: Towards logical quantum computing; Logical qubits: Up to 20; Physical qubits: 1,000+; Gate performance: N/A
2029: Error-corrected computing; Logical qubits: ~100; Physical qubits: 10,000+; Gate performance: 99.9%
2030: Full fault-tolerant scale-up; Logical qubits: ~800; Physical qubits: 100,000+; Gate performance: 99.999999%

By 2029, the company aims to achieve approximately 100 logical qubits with a 99.9% gate performance. The company’s cold-atom laboratory, believed to be among the largest in Israel, is central to these plans. Quantum X Labs intends to integrate its pending patent for a deep transformer decoder, U.S. Patent No. 12,294,387 B2, with its cold-atom control stack to scale from physical to logical qubits.

The company evaluated its decoder on both the publicly available dataset and surface-code simulations to validate its performance. Progress will be measured by a combination of physical qubit count, logical qubit count and gate performance, a methodology the company asserts is more indicative of practical quantum computing power than raw qubit numbers alone. The roadmap’s targets include exceeding 1,000 physical qubits by 2027, and surpassing 10,000 physical qubits by 2029.