In this episode of The New Quantum Era podcast, host Sebastian Hassinger speaks with Steve Girvin, professor of physics at Yale University, about quantum memory - a critical but often overlooked component of quantum computing architecture. This episode was created with support from the American Physical Society and Quantum Circuits, Inc.

Episode Highlights

• Introduction to Quantum Memory: Steve explains that quantum memory is essential for quantum computers, similar to how RAM functions in classical computers. It serves as intermediate storage while the CPU works on other data.
• Coherence Challenges: Quantum bits (qubits) struggle to faithfully hold information for extended periods. Quantum memory faces both bit flips (like classical computers) and phase flips (unique to quantum systems).
• The Fundamental Theorem: Steve notes there’s “no such thing as too much coherence” in quantum computing - longer coherence times are always beneficial.
• Quantum Random Access Memory (QRAM): Unlike classical RAM, QRAM can handle quantum superpositions, allowing it to process multiple addresses simultaneously and create entangled states of addresses and their associated data.
• QRAM Applications: Quantum memory enables state preparation, construction of oracles, and processing of big data in quantum algorithms for machine learning and linear algebra.
• Tree Architecture: QRAM is structured like an upside-down binary tree with routers at each node. The “bucket brigade” approach guides quantum bits through the tree to retrieve data.
• Error Resilience: Surprisingly, the error situation in QRAM is less catastrophic than initially feared. With a million leaf nodes and 0.1% error rate per component, only about 1,000 errors would occur, but the shallow circuit depth (only requiring n hops for n address bits) makes the system more resilient.
• Dual-Rail Approach: Recent work by Danny Weiss demonstrates using dual resonator (dual-rail) qubits where a microwave photon exists in superposition between two boxes, achieving 99.9% fidelity for each hop in the tree.
• Historical Context: Steve draws parallels to early classical computing memory systems developed by von Neumann at Princeton’s IAS, including mercury delay line memory and early fault tolerance concepts.
• Future Outlook: While building quantum memory presents significant challenges, Steve remains optimistic about progress, noting that improving base qubit quality first and then scaling is their preferred approach.

Key Concepts

• Quantum Memory: Storage for quantum information that maintains coherence
• QRAM (Quantum Random Access Memory): Architecture that allows quantum superpositions of addresses to access corresponding data
• Coherence Time: How long a qubit can maintain its quantum state
• Bucket Brigade: Method for routing quantum information through a tree structure
• Dual-Rail Qubits: Encoding quantum information in the presence of a photon in one of two resonators

References

• Weiss, D.K., Puri, S., Girvin, S.M. (2024). “Quantum random access memory architectures using superconducting cavities.” arXiv:2310.08288
• Xu, S., Hann, C.T., Foxman, B., Girvin, S.M., Ding, Y. (2023). “Systems Architecture for Quantum Random Access Memory.” arXiv:2306.03242
• Brock, B., et al. (2024). “Quantum Error Correction of Qudits Beyond Break-even.” arXiv:2409.15065