TRACE: Temporal Residency-Aware Circuit Execution for Over-Capacity Quantum Hardware

Research artifact for the paper accepted at IEEE QCE 2026 (IEEE International Conference on Quantum Computing and Engineering / IEEE Quantum Week), Quantum System Software (QSYS) track. Toronto, Canada, September 13–18, 2026.

Authors: Charles Cao, Sergei V. Kalinin (University of Tennessee, Knoxville); Jong Youl Choi, Ahmad Maroof Karimi (Oak Ridge National Laboratory).

TRACE reframes executing an $N$-qubit circuit on only $M < N$ physical qubits as a temporal qubit residency scheduling problem over the gate stream: it maintains a resident set of $M$ qubits and inserts a wire cut whenever a two-qubit gate references a non-resident operand. Its main practical policy, LR-Gate (Latest-Reuse Gate), scans a bounded window of upcoming gates and evicts the resident qubit whose next two-qubit-gate use lies latest.

This repository ships the code for the paper's simulation experiments and the data that cannot be regenerated:

• src/ contains the scheduler, workload generators, exact oracle, and the drivers for the paper's main simulation experiments. All simulation experiments use fixed random seeds, so their result files regenerate deterministically — no simulation data is shipped.
• data/ibm-hardware/ contains the raw results of the paper's IBM Quantum hardware validation. Hardware runs are not deterministically reproducible (device noise, calibration drift, QPU access required), so this one-run snapshot is included as-is — see the disclaimer in data/ibm-hardware/README.md.

Contents

Path	Role
src/qvm_scheduler.py	TRACE scheduler with the four eviction policies (Random, LRU, LR-Gate, Lookahead)
src/circuits.py	Workload generators (QAOA, QFT, random entanglement, temporal/ancilla-reuse families)
src/temporal_oracle.py	Exact dynamic-programming oracle (OPT) for small instances
src/run_temporal_experiments.py	Order sensitivity (RQ1), oracle gap (RQ2), dynamic-circuit-inspired workloads (RQ4)
src/run_experiments.py	Cut counts, capacity sweep, window ablation, baselines on QAOA/QFT/random workloads
src/run_scalability.py	Scheduler runtime / planning-cost scaling
src/plot_temporal_results.py	Regenerates the temporal-result figures from the JSONs above
data/ibm-hardware/	Single-run IBM Quantum (ibm_fez, 156 qubits) hardware results + disclaimer

Setup

Tested with Python 3.12, Qiskit 2.3.1, qiskit-addon-cutting 0.10.0, networkx 3.6.1, numpy 2.4.4, matplotlib 3.10.8 (matplotlib/seaborn needed only for plotting):

pip install qiskit qiskit-addon-cutting qiskit-aer networkx numpy matplotlib seaborn

Reproducing the simulation results

Each driver writes its JSON outputs into src/results/ (next to the scripts, created on first run — the paths work from any working directory):

python src/run_temporal_experiments.py   # -> src/results/exp_rq1_order_sensitivity.json,
                                         #    src/results/exp_rq2_oracle_gap.json,
                                         #    src/results/exp_rq4_temporal_workloads.json
python src/run_experiments.py            # -> src/results/exp1_* cut-count / ablation / baseline JSONs
python src/run_scalability.py            # -> src/results/exp5_scalability.json
python src/plot_temporal_results.py      # -> figures from the exp_rq* JSONs

Seeds are fixed inside the drivers, so re-running reproduces the numbers reported in the paper for these experiments. The hardware numbers in data/ibm-hardware/ are a one-time snapshot and will not reproduce exactly (see its README).

Citation

@inproceedings{cao2026trace,
  author    = {Cao, Charles and Kalinin, Sergei V. and Choi, Jong Youl and Karimi, Ahmad Maroof},
  title     = {{TRACE}: Temporal Residency-Aware Circuit Execution for Over-Capacity Quantum Hardware},
  booktitle = {IEEE International Conference on Quantum Computing and Engineering (QCE)},
  year      = {2026}
}

License and status

MIT License. This repository is a research artifact archived as published; it is not actively maintained.