Deeper Compilation in the Active Volume Architecture
Athena Caesura
PsiQuantum
Compilation
Resource estimates for the active volume architecture have traditionally been derived by counting the active volume and number of qubits, then substituting these values into a closed-form expression. In this talk, I present software that produces more accurate resource estimates by explicitly scheduling operations. By explicitly assigning qubits to a given task at each logical cycle, the software captures the costs of bridging, stale states, and dynamic memory usage. Using this approach, we achieved a 5× speedup on an EFTQC Fermi–Hubbard simulation task compared with estimates produced by closed-form expressions.
Back to Bridging the Gap Between NISQ and FTQC