Quantum X Labs has activated a 50+ qubit neutral-atom quantum computing platform, marking a step toward scalable quantum processing with a system built around advanced laser cooling and dynamically reconfigurable optical tweezer arrays. The company intends to integrate its patented deep transformer decoder, US12294387B2, an AI-based error correction system, by the end of the first half of 2027, with the goal of reducing computational overhead. This system utilizes syndrome-guided decoding to establish a low-latency feedback loop processing outputs from Rydberg gates in real time. “Our goal is to continue scaling physical qubit counts while building a modular platform with integrated error correction architecture,” said Prof. Nir Sharon, the company’s Chief Quantum Technology Scientist. “This 50+ qubit platform is a vital milestone, giving us the physical environment necessary to implement our proprietary AI-driven error correction and potentially realize efficient, real-time error handling at scale.”
50+ Qubit Neutral-Atom Platform with Laser Cooling Architecture
Quantum X Labs has achieved an advance in neutral-atom quantum computing, bringing a platform with over 50 physical qubits online and demonstrating a scalable architecture for future expansion. This approach facilitates rapid loading of large qubit registers and extends the coherence times crucial for complex calculations, while natively supporting high-performance Rydberg-mediated two-qubit gates. Central to the company’s strategy is a pathway toward logical qubits, aiming to reach thousands of qubits by the end of H1 2027. This real-time processing of Rydberg gate outputs is designed to reduce computational overhead, enabling practical, fault-tolerant logical qubit operations. The newly launched platform is a functioning system intended to support applications across diverse fields, including aerospace and defense, clinical trials, and quantum cybersecurity.
Deep Transformer Decoder for Low-Latency Error Correction
The system’s error mitigation is deeply interwoven with the core control mechanisms of the quantum processor, rather than being an add-on. This real-time processing is designed to reduce the computational overhead traditionally associated with error correction, a significant hurdle in scaling quantum systems. The system’s architecture aims for a pathway toward efficient, real-time error handling at scale. This hardware-AI co-design represents a departure from conventional error correction methods, which often rely on extensive post-processing. By embedding the deep transformer decoder directly within the control stack, Quantum X Labs hopes to achieve a level of responsiveness necessary for complex quantum computations.
This 50+ qubit platform is a vital milestone. It gives us the physical environment necessary to implement our proprietary AI-driven error correction and potentially realize efficient, real-time error handling at scale.
The company’s approach to error mitigation is deeply interwoven with the platform’s core functionality. This hardware-AI co-design is intended to enable practical, real-time fault tolerance in logical qubit operations. Quantum X Labs’ comprehensive portfolio extends beyond quantum computing itself, encompassing quantum software, simulation, and sensing technologies.
Our goal is to continue scaling physical qubit counts while building a modular platform with integrated error correction architecture.
