Key Takeaways: Chapter 28 — Photonic, Neutral Atom, and Other Approaches: The Diversity of Quantum Hardware and Why No One Has Won Yet
-
Photonic quantum computing uses single photons as qubits, with gates mediated by linear optics and measurement. The KLM scheme proves universality, while boson sampling provides a near-term quantum advantage demonstration. The fundamental challenge is that photons don't interact with each other — nonlinearity must be manufactured.
-
Neutral atom qubits trapped in optical tweezer arrays offer the largest qubit counts (1000+) and longest coherence times (40+ s for Sr), with Rydberg blockade enabling flexible, high-connectivity gates. Quantum is linear algebra — the atom choice determines the qubit quality, and alkaline-earth atoms provide naturally decoupled nuclear spin qubits.
-
Silicon spin qubits promise CMOS-compatible fabrication and extreme scalability, but currently lag in gate fidelity and qubit count. The path to millions of qubits runs through semiconductor foundries, but noise is the enemy — charge noise, valley splitting, and nuclear spin noise must be controlled.
-
Topological qubits based on Majorana zero modes offer the theoretical promise of hardware-level error protection, but no operational qubit has been demonstrated. The topological protection comes from encoding information non-locally — a fundamentally different approach to fighting errors.
-
NV centers in diamond operate at room temperature with millisecond coherence, making them attractive for sensing and small-scale computing. Their niche is quantum sensing rather than large-scale computing.
-
No single platform dominates all figures of merit. The future likely involves heterogeneous architectures combining the strengths of multiple platforms. Quantum advantage is problem-specific — the best platform depends on the application. We're at the beginning — the hardware landscape will continue to evolve rapidly.