A team of scientists from Caltech, led by graduate students Alkim Bozkurt and Omid Golami and supervised by Mohammad Mirhosseini, has developed a hybrid approach for quantum memories. The new work, detailed in a paper published in Nature Physics, demonstrates a system that translates electrical information into sound to store quantum states from superconducting qubits for a period up to 30 times longer than other techniques.
The researchers fabricated a superconducting qubit on a chip and connected it to a mechanical oscillator, a miniature tuning fork device that vibrates with sound waves at gigahertz frequencies. When an electric charge is applied to the oscillator’s plates, it can interact with electrical signals carrying quantum information, allowing information to be stored and retrieved. The use of acoustic waves offers several advantages, as they travel slower than electromagnetic waves, enabling more compact devices, and do not propagate in free space, which helps prevent energy from leaking out of the system.
This method offers a potential pathway for quantum memories, which are necessary for storing quantum states for later use in logical operations, a capability where superconducting qubits have traditionally faced limitations. The longer lifetime of the mechanical oscillators addresses a challenge in these systems. The advantages of this platform point to a potentially scalable method for creating quantum memories by including many such oscillators on a single chip.
Read the full announcement here and the paper in Nature Physics here.
August 15, 2025
