Quantum In-Memory Computing

Researcher:
Associate Prof. Yachin Ivry | Materials Science and Engineering

Categories:

Computer Science & Electrical Engineering | Electro-Optics & Quantum

The Technology

This invention is based on Hybrid Superconducting-Ferroelectric Quantum Memristors. The novel platform integrates memory capability directly into a quantum superconducting device — combining the strengths of a memristor with those of a quantum superconducting circuit. At its core, the design replaces the metal electrodes of a Ferroelectric tunnel junction with superconductors creating a Josephson junction with a ferroelectric barrier, establishing a quantum memristor. In such hybrid quantum devices with memory characteristics a ferroelectric layer serves as the tunneling barrier in a superconducting Josephson junction; thus, the potential barrier is history-dependent. This yields in a Memory-Integrated Qubit, paving the way for current-biased self-modulating quantum devices with built-in memory — merging computing and memory functions in the quantum domain.

Advantages

• Built-in non-volatile memory: Since the ferroelectric polarization persists without power, the quantum memristor retains its internal classical and quantum state — enabling long-lived memory in a quantum circuit.
• Integration of memory + computing: By combining memory and quantum logic in one physical junction, this technology reduces the need for separate memory and qubit hardware, potentially lowering device complexity, footprint, and interconnect overhead.
• History-dependent behavior (self-adaptation): The junction’s tunneling barrier — and thus its resistance / critical current — depends on past currents or voltages, enabling adaptive behavior and possibly neuromorphic-inspired quantum devices.
• Energy efficiency potential: With memory and computing co-located, data transfer overhead between separate memory and computing units is significantly reduced — improving energy efficiency, especially for memory-intensive quantum tasks or neuromorphic-quantum architectures.
• Reduced error rate: Modulating the potential energy of the qubit through the coupling with the ferroelectric polarization can improve error rates in quantum computing.

Applications & Opportunities

• Quantum computing and integration with classic computing systems
• Neuromorphic and AI Quantum applications
• Low-energy and low-complexity quantum computing architectures

arrow Business Development Contacts
Dr. Arkadiy Morgenshtein
Director of Business Development, ICT
our new tracks