The big picture and the nitty-gritty: A trip to IWQC 2025

Imagine a conference that feels less like a series of formal lectures and more like an intense, collaborative jam session. That was the energy at the 7th International Workshop on Quantum Compilation (IWQC) in Helsinki. The conversations—fueled by plenty of coffee—revealed some major shifts in the quantum community.

The main takeaway? Discussions emphasized that the future of quantum compilation is moving beyond simple optimizations. It's about building a collaborative, open-source ecosystem, creating shared benchmarks, and developing new tooling to allow different compilers to work together.

Weren't able to attend? Worry not—read on to get a front-row seat to the most important insights from this event.

Seeing the big picture: compilation beyond the gate model

As mentioned above, an important theme and overarching question for modern quantum compilation research is:

How do we optimize quantum programs at higher abstraction levels than gates?

Olivia had the honour of opening the workshop with a keynote posing this question and established higher-level optimization as a key theme of IWQC 2025. She highlighted the challenges and opportunities of structured quantum programs and non-local circuit optimization, using the concrete example of constant-time compilation of Shor's algorithm. This message was reinforced through the presentations by Mark Koch (Quantinuum), using affine relational analysis and Pauli tracking for non-local optimizations of structured programs, and Vivien Vandaele (Quantinuum), who showed how to lower the computational overhead of Pauli product rotation compression by looking at the full picture of their commutation graph.

Later talks also emphasized that the "big picture" approach to quantum compilation has even more to offer. In the second keynote, Adam Glos (Algorithmiq), presented a software framework for compilation of fermionic quantum circuits, which applies optimizations to the compilation pipeline itself to reduce computational cost. An entirely different take was presented by Marcin Szyniszewski (University of Oxford), who showcased improved rotation gate approximations by replacing single quantum circuits by classical mixtures, which leads to genuinely hybrid programs. This talk, which closed out the workshop, was followed by a classical mixture of the attendees.

Sibelius monument in Sibelius park in Töölö, Helsinki.

Pull out the magnifying glass, we’re about to get local

The theme of high-level optimization was complemented by a series of talks on decomposition methods. Despite decades of research in the area, we are always impressed to see novel circuit designs and improved resource counts, be it multi-qutrit gates for graph colouring applications (Daniele Trisciani, Aalto University), or quantum adders without auxiliary qubits that achieve (poly)logarithmic depth (Vivien Vandaele, Quantinuum). David was happy to contribute a talk on recursive Cartan decompositions for unitary synthesis, a framework that allows for a systematic analysis and extension of synthesis techniques, but also for compilation in practical applications such as fast-forwarded time evolution.

Tanuj Khattar (Google) showed how to compile modular gcd computations for decoded quantum interferometry, a recently proposed protocol to achieve quantum speedups for certain classical optimization tasks. Yusei Mori (Osaka University) showcased a new multi-product commutation relation between Pauli rotation gates, discovered while on a quest for unoptimization techniques of discrete circuits that could challenge existing compilers. Another great example of synergy, this time between practical benchmarking efforts and compiler research.

What about errors?

While many other talks touched on bridging and interactions between different abstraction levels, perspectives or tools, we noticed that physical compilation that takes error correction into account is yet to be fully integrated into the stack. Building closer connections between the fundamental building blocks of a fault-tolerant quantum computer and the more established logical compilation layers remains a challenging task, and calls for effort, expertise, and creativity from across the community.

Multiple talks showcased how to move into this direction; Eddie Schoute (IBM) presented recently developed bivariate bicycle error correction codes, complemented by hardware experiments to prototype their implementation. Erik Weilandt (TUM) showcased practical compilation of error-detected state preparation, and Ali Javadi-Abhari presented spacetime codes for error detection.

Open up black boxes—and closed sources!

We already anticipated it at the top, but want to reiterate another interesting topical focus at IWQC, maybe best represented by the following observation we took away:

The quantum compilation community thrives on communication and open source software projects.

This resonated with us and our interest in putting compilation concepts to work in practice, and showed that the field is constantly at work building bridges between abstract concepts and software tooling. While this message could be heard throughout the workshop, it was made explicit by Jordan Sullivan (Unitary Foundation, UF) during their talk on the Unitary Compiler Collection (ucc).

UF’s community-driven compiler aims for nothing less than becoming the “gcc for quantum.” The topic was also represented particularly strongly in the poster session, for example through connectivity-aware transpilation, hardware-level optimization, and compilers for distributed quantum computing. Yannick Stade (TUM) complemented this with a talk on a routing-aware neutral atom placement.

Most of these projects came with software implementations, or even presented them as their main result, and Maxime Garnier (Inria) presented Graphix, an open-source compiler and simulator for measurement-based quantum computation (MBQC). Brad Chase (Unitary Foundation) presented one particular component of the Unitary Compiler Collection, ucc-ft, which enables users to perform formal verification of fault-tolerant quantum circuits.

The collaborative spirit of the quantum compilation community was on full display at IWQC 2025. This wasn't just a conference; it was a vibrant, two-day exchange of ideas that reinforced the importance of open communication and cooperation. Many thanks to the local organizers, Alexandru Paler (Aalto University) and Arianne Meijer-van de Griend (University of Helsinki)!

Foggy coastline in Helsinki as seen on our way to the conference dinner.

From the engaging talks to the lively discussions, IWQC 2025 was a fantastic experience that highlighted the need for open-source solutions and shared benchmarks—and open-source software is the key to accelerating progress in quantum compilation.

Looking to get involved? We invite you to explore and contribute to the PennyLane Compilation hub, where you can find tools, resources, and a community dedicated to advancing quantum compilation.