Scalable Passive QRAM

The most promising candidates for practical quantum advantage lie firmly in the "small-data” regime. This is caused by the data loading issue, which dominates the cost of running quantum algorithms on big-data problems. Classical algorithms can access massive datasets cheaply because memory access is passive. Conventional quantum data access, by contrast, is inherently active and incurs an energy cost that grows linearly with input size.

In this talk, I will present how to build a QRAM chip which consumes O(log N) energy per query over a size-N dataset, and runs in O(log^2 N) time.

Based on joint work with Alex Dalzell & Connor Hann: https://scirate.com/arxiv/2610.03700.