Automatic Resource Estimation via Efficient Compilation of Key Quantum Primitives

As quantum hardware approaches thresholds for fault tolerant computation, there remains an important role for NISQ compilation and experiments. Our resource estimation tool represents and important bridge between the NISQ and Fault Tolerant eras of quantum computation. Many resource estimates suffer from being either too specific to a particular application or architecture or from being too general without enough focus on the implementation details. This imbalance creates a challenge for many hardware designers to know which major architectural improvements will be the most important over different time scales. Our work frames resource estimation as a compilation problem, where compiled fault tolerant circuits rely on a handful of primitive operations that may be run or simulated on NISQ devices. These small logical operations can be intensely omptimized in the NISQ setting to minimize the resources needed to run in the fault tolerant regime, providing important stepping stones and benchmarks for NISQ compilation and experiments. Through this feedback loop, we expect to be able to isolate key hardware design decisions for different applications that will have the most benefit for the cost of implementation. In the space of neutral atoms in particular, where there are many options for architectures being explored simultaneously, software to automatically compare the impacts of these decisions at the application level will be particularly useful. In this talk we demonstrate the impact of work through a comparison of several different examples of neutral atom architectures and applications.