A compilation-driven framework accurately estimates the resources required for computation, addressing a key hurdle in realising practical fault-tolerant quantum computers. Colin Campbell and colleagues at University of Chicago bridge the gap between hardware design and circuit implementation, offering a more nuanced approach than existing methods. The framework translates quantum circuits into fundamental operations with defined physical costs, enabling rapid assessment of different architectural choices, particularly for neutral atom quantum computers. Applying this to quantum simulation and optimisation tasks using the surface code, the researchers identify key architectural trends, notably the importance of efficient qubit movement and frugal routing as problem sizes increase, suggesting dual-species arrays with controlled movement could pave the way for near-term quantum advantage. Qubit movement unlocks substantial gains in neutral atom quantum circuit compilation Access to movement within neutral atom quantum computers reduces circuit compilation time by a factor of two, a feat previously unattainable without significant architectural redesign. Neutral atom qubits offer a promising platform for quantum computation due to their long coherence times and all-to-all connectivity potential, but realising this potential requires overcoming challenges related to qubit control and routing. Previously, accurately modelling the interaction between qubit routing and gate overhead proved impossible due to the complexity of simulating realistic hardware constraints, but this framework now enables that assessment. The core innovation lies in a compilation pipeline that explicitly accounts for the physical costs associated with qubit movement, gate operations, and measurement. This allows for a more holistic evaluation of architectural trade-offs, moving beyond simplified analytical models. Translating quantum circuits into logical primitive operations with defined physical costs rapidly assesses architectural choices, revealing that routing and qubit movement become dominant bottlenecks as problem size increases; this highlights the potential of dual-species arrays with controlled qubit movement for near-term quantum advantage. The framework decomposes complex quantum