QASMTrans: An End-to-End QASM Compilation Framework with Pulse Generation for Near-Term Quantum Devices

QASMTrans is a lightweight, high-performance, C++-based quantum compiler that bridges abstract quantum algorithms to device-level control and is designed for just-in-time (JIT) deployment on QPU testbeds with tightly integrated FPGAs or CPUs. We focus on achieving fast transpilation times on circuits of interest, we find more than 100x faster compilation than Qiskit in some circuits with similar circuit quality, enabling transpilation of large, high-depth circuits in seconds. Unlike existing tools, QASMTrans offers end-to-end device-pulse compilation and direct quantum control integration with QICK, closing the gap between logical circuits and hardware control enabling closed-loop optimization. QASMTrans supports latency-aware Application-tailored Gate Sets (AGS) at the pulse level, identifying high-impact gate sequences on the circuit critical path and synthesizing optimized pulse schedules using pre-defined robust circuit ansatz. Validated through integrated QuTiP pulse-level simulation, this is found to significantly reduce execution latency and can improve final-state fidelity by up to 12% in some tested circuits. QASMTrans further implements device-aware, noise-adaptive transpilation that uses device calibration data for circuit placement on high-quality qubits and can focus on the circuit critical path to reduce transpilation-pass time while maintaining comparable fidelity. Additionally, it introduces circuit space sharing via calibration-aware device partitioning, enabling concurrent execution of multiple circuits or shots on a single QPU. Moreover, QASMTrans is entirely self-contained and has no external library dependencies, making it easy for practical deployment. By combining fast compilation, pulse-level control, and noise-aware optimization, QASMTrans enables real-time adaptive algorithms such as ADAPT-VQE and ADAPT-QAOA.

💡 Research Summary

QASMTrans is a lightweight, high‑performance quantum compiler written entirely in C++ that bridges the gap between abstract quantum algorithms expressed in OpenQASM and the low‑level pulse control required by near‑term quantum processors. The framework is organized into four principal modules: (1) an I/O subsystem that parses OpenQASM 2.0 files into an internal intermediate representation (IR) and reads device specifications (topology, basis‑gate set, calibration data) from JSON; (2) a configuration stage that decomposes any three‑qubit gates into one‑ and two‑qubit primitives, constructs the device coupling graph, and builds a directed‑acyclic graph (DAG) of the circuit; (3) an optimization engine that goes beyond conventional gate‑decomposition, routing, and mapping by introducing two novel strategies. First, it performs latency‑aware Application‑tailored Gate Set (AGS) synthesis: the compiler analyses the critical path of the circuit, identifies high‑frequency sub‑sequences, and replaces them with pre‑designed robust pulse‑level ansätze, thereby reducing SWAP overhead and overall execution latency. Second, it implements noise‑adaptive, calibration‑aware placement: qubits with the longest T1/T2 times and lowest gate error rates are preferentially selected, and dynamic remapping is applied to keep the critical path short. The optimizer also supports calibration‑aware device partitioning, enabling multiple circuits or shots to be executed concurrently on a single QPU, effectively increasing throughput. (4) The final transpiler component emits the transformed QASM and a schedule of microwave pulses, which are handed directly to the QICK (Quantum Instrumentation Control Kit) interface for immediate deployment on FPGA‑ or CPU‑based control hardware. Because QASMTrans has no external library dependencies, it can be compiled for embedded ARM platforms as well as for leadership‑class HPC systems, offering broad portability.

Performance evaluation was conducted on four major quantum platforms—IBM superconducting, Rigetti superconducting, IonQ trapped‑ion, and Quantinuum trapped‑ion—using the QASMBench suite (circuits ranging from 4 to 127 qubits and from 10 k to 500 k gates). Compared with Qiskit’s O3 optimization level, QASMTrans achieved compilation speedups of 30× to over 100×, often completing large, deep circuits in a few seconds where Qiskit required minutes. Circuit quality, measured by depth, gate count, and simulated fidelity, remained comparable to Qiskit; in many cases the depth was slightly reduced. Pulse‑level simulations performed with a custom QuTiP backend showed that the AGS‑derived pulse schedules could improve final‑state fidelity by up to 12 % relative to standard gate‑level compilation. Real‑hardware experiments confirmed that the fidelity deviation from Qiskit was under 1 % across all platforms, demonstrating that the speed gains do not come at the expense of accuracy. The calibration‑aware partitioning feature allowed two or three circuits to share a single QPU without significant crosstalk, yielding a 2–3× increase in overall throughput.

The paper’s contributions are: (i) a fully self‑contained C++ quantum transpiler that dramatically reduces compilation latency; (ii) latency‑aware AGS synthesis that integrates pulse‑level optimization directly into the compilation flow; (iii) noise‑adaptive placement and device partitioning that exploit real‑time calibration data; (iv) seamless integration with QICK for just‑in‑time, closed‑loop execution, enabling adaptive algorithms such as ADAPT‑VQE and ADAPT‑QAOA; and (v) extensive cross‑platform validation showing comparable or superior circuit quality with orders‑of‑magnitude faster compilation. Future work will explore richer pulse‑ansatz libraries, machine‑learning‑driven mapping heuristics, and integration with quantum error‑correction protocols.