MIT and MITRE researchers have created a “quantum-system-on-chip” (QSoC) that integrates thousands of qubits, enabling scalable and precise quantum computing.
Credits:Image: Sampson Wilcox and Linsen Li, RLE
Quantum computers could swiftly solve complex problems that take current supercomputers decades, but building a system with millions of interconnected qubits presents a significant global challenge.
Researchers at MIT and MITRE have developed a “quantum-system-on-chip” (QSoC) that integrates thousands of qubits for scalable, precise quantum computing. This system supports “entanglement multiplexing” and links multiple chips via optical networking to form a quantum network. The team has also perfected a method to efficiently transfer two-dimensional qubit arrays onto a CMOS chip in one step.
Diamond microchip lets
The researchers chose diamond colour centres as qubits for their scalability and previous integration with photonic circuitry. These qubits, featuring photonic interfaces, allow remote entanglement with non-adjacent qubits, which is essential for managing inhomogeneity in large systems. They integrated an extensive array of these qubits onto a CMOS chip with digital logic that adjusts voltages automatically, ensuring connectivity across thousands of qubits.
Lock-and-release fabrication
To develop the QSoC, researchers devised a large-scale method to transfer diamond colour centre “microchiplets” onto a CMOS backplane. Starting with an array from a diamond block and designing nanoscale antennas to collect photons, they post-processed a CMOS chip to align with these microchiplets. Using a lock-and-release technique in-house, they secured the microchip, letting in the chip sockets, enabling the transfer of a 500-micron area with the potential for larger-scale applications. Scaling up also reduces the voltage needed for frequency tuning.
Following extensive nanostructure testing for optimal integration, the team devised methods to measure system performance extensively. They used a custom cryo-optical metrology setup to tune over 4,000 qubits to the same frequency while maintaining their properties. Additionally, they developed a digital twin simulation to enhance understanding and implementation of the architecture, connecting experimental results with digital models for better insight into the quantum system’s functionality.
In the future, researchers could enhance their system by improving qubit materials or refining control processes. They might also extend this architecture to other solid-state quantum systems.
