IBM Connects Two Nighthawk Quantum Processors: 99% Qubit Transmission Accuracy Paves the Way for Modular Supercomputers

For the first time in the history of quantum computing, IBM engineers have demonstrated stable quantum information transfer between two independent processors with an accuracy of approximately 99%. The experiment, reported by 3DNews, marks a key milestone toward modular quantum computers—systems where multiple separate chips work as a single entity. Essentially, this brings the industry closer to creating quantum supercomputers capable of performing tasks beyond the reach of any single die.

Two Nighthawks, Four Cables, 80 Centimeters

The experiment was based on two IBM Quantum Nighthawk superconducting processors, each containing 120 qubits. These chips were announced about a year prior to the reported test. To couple them, engineers used four 80 cm aluminum coaxial cables operating at microwave frequencies of around 5 GHz. Each cable formed a separate communication channel between specialized tunable qubits located at the ends of the connection. The cables were installed perpendicularly to the chip surfaces, and the contacts were secured using indium-indium pressure joints at room temperature. Notably, the assembly was performed after the processor modules had already been installed in the cryogenic system, significantly simplifying maintenance and scalability of the setup. Meanwhile, only a subset of qubits from each processor was connected and measured for the experiment, allowing for future expansion of active channels.

Entanglement and Cross-Chip Two-Qubit Operations

Beyond simple quantum state transfer, the researchers demonstrated the joint execution of two-qubit operations and the creation of entangled states, including simultaneously across two adjacent channels. The average fidelity of inter-chip CZ operations—controlled-Z, a two-qubit quantum operation linking qubits on different dies—was 94.75%, with the best recorded result reaching 96.42%. These figures are slightly lower than state transfer fidelity because the operation itself involves several additional quantum transformations. Crucially, all four connections remained operational after three full thermal heating and cooling cycles without contact failure, confirming the mechanical and electrical reliability of the proposed architecture.

L-coupler: Proven Connector Functionality

The experiment clearly showed that the L-couplers introduced alongside the Nighthawk processors are capable of linking independent quantum processors and transferring quantum information between them. This is not merely a laboratory curiosity, but proof that scaling quantum computers is technically feasible. The ability to integrate chips into modules opens the path to increased performance and fault tolerance in quantum computers, enabling more complex algorithms and efficient error correction codes that are critical for moving from demonstration experiments to practical computing.

"Starling": The First Modular Quantum Supercomputer by 2029

IBM has already developed standard cryogenic racks for modular systems, and the company plans to use them to build the first practically viable modular quantum supercomputer, codenamed Starling. The successful experiment connecting two quantum processors confirmed that quantum data can be transmitted between adjacent racks and further down the chain. According to the company, IBM now has a clear roadmap to achieve its goal by 2029, in line with its previously announced quantum development plan.

Industry Significance

IBM's achievement fits into the broader industry trend of transitioning from single chips to modular architectures. While the primary limitation of quantum computers was previously the number of qubits on a single die, the focus is now shifting to the quality of inter-chip links and long-distance entanglement stability. A transmission accuracy of around 99% and the retention of functionality after thermal cycles are the parameters without which no modular quantum computer can leave the research lab. The Nighthawk experiment demonstrates that the engineering foundation for this transition is already in place, and the remaining challenges are primarily systemic rather than fundamental.