The operating environment for live quantum computation — managing logical qubit pools, T-state reserves, and decoherence monitoring as a coordinated, persistent system that keeps a device running safely over extended periods.
A quantum device is not a simple request-response system. Qubits are consumed and released, T-state factories must be replenished, and decoherence accumulates continuously. Left unmanaged, these competing demands degrade computation quality silently until results become unreliable. The persistent quantum runtime monitors and coordinates all three concerns in real time.
The runtime operates through three coordinated subsystems: CoreFire manages the pool of active logical qubits; FuelFire maintains the T-state reservoir needed for fault-tolerant gate operations; and GuardFire continuously monitors decoherence rates and triggers protective action when quality drops below threshold. The threshold is set at 1/φ — the golden ratio reciprocal — which the framework identifies as the natural boundary for profitable computation. If GuardFire determines that the current decoherence level makes continued computation uneconomical, the runtime gracefully suspends rather than producing corrupted results.
CoreFire tracks which logical qubits are in use, allocates and releases them for circuit execution, and maintains their error-correction state between circuit runs. The pool is available to the fault-tolerant executor without requiring per-circuit initialization overhead.
FuelFire manages the supply of T states needed by fault-tolerant gates, monitoring factory throughput and signaling when supply is insufficient for scheduled operations. Shortfalls are surfaced before execution begins, not during it.
GuardFire samples decoherence rates in real time and compares them against the 1/φ profitability threshold, triggering graceful suspension before quality degrades past recovery. This prevents silent result corruption during extended computation sessions.
Rather than continuing execution through severe decoherence events, the runtime suspends computation at a clean checkpoint and waits for conditions to recover, preserving what work has been done and allowing the session to resume without data loss.
Operations layer — the runtime is the substrate on which live quantum computation runs. The fault-tolerant executor draws qubit and T-state resources from it; the coherence recycler surfaces coherence-degraded outputs to it for recovery assessment; GuardFire feeds the decoherence monitor data that informs whether ZNE or recycling interventions are warranted.
Draws qubit and T-state resources directly from the runtime, relying on CoreFire and FuelFire to supply what each circuit run requires.
GuardFire's decoherence monitor surfaces candidates for coherence recycling, feeding degraded outputs to the recycler for recovery assessment.
Published under the GNU AGPLv3 with whitepaper and reference implementation. Commercial licensing is available for closed-source deployments.