The MQT Compiler Collection: A Blueprint for a Future-Proof Quantum-Classical Compilation Framework
Lukas Burgholzer, Daniel Haag, Yannick Stade, Damian Rovara, Patrick Hopf, Robert Wille
Abstract
As the capabilities of quantum computing hardware continue to rise, algorithms that exploit them are becoming increasingly complex. These developments increase the need for sophisticated compilation frameworks that translate high-level algorithms into executable code. In the past, most solutions were built with a quantum-first approach and handled mostly pure quantum programs without classical elements such as structured control flow. However, developments in quantum algorithms, error correction, and optimization, as well as the integration into high-performance computing (HPC) environments, depend on such classical elements. As quantum-first approaches increasingly struggle to handle these concepts, classical-first approaches are becoming a promising alternative. In this work, we present the MQT Compiler Collection, a blueprint for a future-proof quantum-classical compilation framework built on the Multi-Level Intermediate Representation (MLIR). After years of experience with the quantum-first approach and its shortcomings, we propose a framework that embraces core MLIR concepts to support the full compilation pipeline from high-level algorithms to hardware-specific instructions. The proposed architecture is designed from the ground up to support complex optimizations beyond, e.g., simple gate cancellation. It is publicly available at https://github.com/munich-quantum-toolkit/core.
AI Impact Assessments
(3 models)Scientific Impact Assessment: The MQT Compiler Collection
1. Core Contribution
The paper presents the MQT Compiler Collection (mqt-cc), a quantum-classical compilation framework built on MLIR (Multi-Level Intermediate Representation). The central architectural idea is a dual-dialect design: an imperative QC dialect for I/O and hardware interfacing (using reference semantics), and a functional QCO dialect for optimization (using value semantics/linear types). The framework supports progressive lowering from high-level quantum algorithms to hardware-specific instructions, with bidirectional conversion between the two dialects.
The key intellectual contribution is the articulation of a "classical-first" design philosophy—treating quantum hardware as an accelerator (akin to a GPU) within a broader classical compilation ecosystem, rather than building quantum-specific infrastructure from scratch. This is positioned as a synthesis of lessons learned from years of quantum-first development (the Munich Quantum Toolkit) with the proven methodology of MLIR.
2. Methodological Rigor
This is fundamentally an architecture/systems paper rather than an empirical research contribution. As such, it should be evaluated on the soundness of its design choices and the quality of its implementation, rather than on experimental validation.
Strengths in rigor:
Weaknesses in rigor:
3. Potential Impact
The potential impact operates on several levels:
Community resource: As an open-source MLIR-based quantum compilation framework, this could serve as a reference implementation for the community. The quantum compilation ecosystem is fragmented, and a well-designed MLIR-based framework could become a unifying substrate.
HPC integration: The classical-first approach is particularly relevant as quantum computing moves toward integration with HPC environments. By treating quantum operations as "just another dialect," the framework naturally supports heterogeneous compilation pipelines.
Extensibility: The modular architecture (demonstrated by the PennyLane plugin integration) suggests good extensibility, which could attract external contributors.
Limitations on impact:
4. Timeliness & Relevance
The paper addresses a genuine and timely need. The quantum computing community is indeed at an inflection point where:
The adoption of MLIR for quantum compilation is an emerging trend (Catalyst, CUDA-Q's Quake dialect), and this paper contributes to that trend. However, it is not the first to propose MLIR for quantum compilation—it arrives in a space already populated by several similar efforts, which somewhat diminishes its novelty.
5. Strengths & Limitations
Key Strengths:
Notable Limitations:
Overall Assessment
This paper presents a thoughtful architectural vision for quantum-classical compilation on MLIR, backed by a concrete open-source implementation. The dual-dialect design is the most distinctive technical contribution. However, the paper reads more as a position/vision paper than a research contribution with validated claims. The complete absence of empirical evaluation—no benchmarks, no comparisons, no scalability analysis—significantly limits the ability to assess whether this architecture delivers on its promises. The impact is primarily as a community resource and architectural reference, rather than as a demonstrated advance in compilation quality or capability.
Generated Apr 13, 2026
Comparison History (40)
Paper 2 offers a broader and more immediate scientific impact due to its highly practical application in quantum computing infrastructure. As quantum hardware scales, integrating quantum and classical workflows is a critical bottleneck. By providing a scalable, open-source MLIR-based compilation framework, Paper 2 enables advancements across algorithms, error correction, and HPC integration, potentially becoming a widely-adopted foundational tool. In contrast, while Paper 1 presents significant theoretical advances in quantum statistical mechanics, its immediate impact is more confined to the specific subfield of quantum many-body systems and ensemble preparation.
Paper 2 is likely to have higher scientific impact: it introduces a systematic, potentially broadly applicable bootstrap methodology for infinite-lattice open quantum many-body systems with absorbing phase transitions, yielding quantitative bounds on critical properties and spectral gaps. This is a novel theoretical tool with cross-field relevance (nonequilibrium statistical physics, quantum information, many-body theory) and strong timeliness given growing interest in dissipative quantum matter and Lindbladian dynamics. Paper 1 is valuable infrastructure, but compiler frameworks are crowded and impact depends heavily on long-term adoption and ecosystem effects.
Paper 1 likely has higher impact due to broad, timely applicability: a future-proof quantum-classical compilation framework built on MLIR can influence many quantum software stacks, enable integration with HPC, and support diverse algorithms and error-correction workflows. It also delivers an open-source toolchain, increasing adoption and downstream research. Paper 2 is novel and potentially high-impact within quantum algorithms for nonlinear higher-order dynamics, but its assumptions and narrower domain may limit near-term real-world uptake compared with a general-purpose compiler infrastructure.
Paper 2 likely has higher impact: it proposes an end-to-end, quantitatively evaluated fault-tolerant trapped-ion architecture based on LDPC codes, including compiler, micro-architecture, decoder, and simulations with concrete resource and runtime estimates toward classically intractable physics simulations. This directly targets a central bottleneck (practical fault tolerance) with near-term experimental relevance and broad implications across quantum hardware, QEC, compilation, and applications. Paper 1 is valuable infrastructure (MLIR-based quantum-classical compilation) with good applicability, but its scientific impact is more incremental/tooling-focused versus Paper 2’s system-level blueprint and performance claims.
Paper 2 likely has higher impact due to broad, immediate applicability: a public, MLIR-based quantum-classical compiler framework can affect many algorithms, hardware backends, and HPC workflows, and can become infrastructure used across academia/industry. Its timeliness is high as hybrid quantum-classical execution, structured control flow, and integration with error correction/optimization are pressing needs. Paper 1 is novel and potentially important, but relies on amplitude estimation (often deep/coherence-heavy) which may limit near-term practicality and adoption; impact is narrower to QNN inference/training scenarios.
Paper 1 addresses a critical near-term bottleneck in quantum computing by proposing a scalable, MLIR-based compiler framework for quantum-classical algorithms. Its practical utility, open-source availability, and relevance to integrating quantum computing with high-performance computing (HPC) give it a higher potential for broad, immediate impact across the rapidly growing quantum software and hardware communities compared to the foundational, yet narrowly theoretical physics focus of Paper 2.
Paper 2 proposes foundational software infrastructure for hybrid quantum-classical computing. As quantum hardware scales, integrating quantum algorithms with classical control flow and high-performance computing (HPC) is a critical bottleneck. A robust, open-source compiler framework based on MLIR will likely be adopted by a wide audience of algorithm developers and hardware researchers, leading to broad, cross-disciplinary impact. Paper 1 presents a highly rigorous and innovative numerical method for computational quantum physics, but its impact is more niche and confined to specific phase-space simulation domains.
Paper 2 addresses a fundamental and broadly impactful challenge in quantum computing—building a future-proof compilation framework that bridges high-level quantum algorithms and hardware execution using MLIR. This has wide applicability across quantum algorithms, error correction, and HPC integration, serving a large and growing community. Paper 1, while technically rigorous in characterizing pump phase noise in JTWPAs, addresses a more specialized hardware concern with narrower impact. Paper 2's open-source framework and architectural blueprint position it to influence the entire quantum software stack as the field scales.
Paper 2 is more scientifically novel: it introduces thermodynamic, measurement-light witnesses for quantum “magic,” linking resource theory to experimentally accessible energy/heat observables. This can impact multiple areas (quantum thermodynamics, many-body physics, certification/verification, fault-tolerance resource quantification) and is timely for near-term experiments lacking full tomography. Paper 1 is valuable engineering infrastructure, but its core idea (MLIR-based compilation framework) is less conceptually novel and its impact depends on adoption and implementation details, with narrower cross-field scientific reach.
Paper 2 likely has higher impact due to broad, immediate applicability: a publicly available MLIR-based quantum-classical compilation framework can influence many researchers and industrial workflows across quantum software, HPC integration, optimization, and error-correction toolchains. Its “classical-first” architecture addresses a timely bottleneck as hybrid programs and structured control flow become central. Paper 1 is scientifically novel and rigorous with a strong experimental result in multi-parameter quantum metrology, but its near-term user base and cross-field spillover are narrower than a foundational compiler infrastructure adopted by the community.