Quantum Resource Estimation Workshop

Resource Estimation Driven Optimization of Quantum Phase Estimation for Chemistry Applications
David Williams-Young, Microsoft

Abstract: The practical cost of the Sum-of-Squares Spectral Amplification (SOSSA) algorithm applied to quantum phase estimation for electronic structure is governed as much by its classical preprocessing as by its quantum circuitry. We report recent Microsoft Quantum work in which resource estimates taken over explicit circuit implementations, rather than phenomenological cost models, serve as the optimization target for both. Running this loop end to end lets us optimize the algorithm for architectures of interest, and to assess how the preferred implementation shifts as those architectural assumptions change. We also discuss the implementation of this pipeline in the Microsoft Quantum Development Kit (QDK) through which the resulting workflows are exposed as modular, reproducible components from molecular specification through to quantum resource estimation.

Dr. David Williams-Young is a Principal Quantum Software Architect at Microsoft Quantum. His work focuses on quantum computing applications in chemistry and materials science, including quantum algorithms, classical simulation methods, and the development of tools that bridge quantum computing and computational many-body theory. Prior to joining Microsoft, he was a Scientist in the Applied Mathematics and Computational Research Division at Lawrence Berkeley National Laboratory, where he developed exascale electronic structure methods and software for DOE Leadership Computing Facilities. He received his Ph.D. in Chemistry from the University of Washington, specializing in relativistic electronic structure theory. He is the author of numerous open-source computational chemistry libraries and has served as a major contributor to numerous quantum chemistry software packages.

Accelerating Quantum Algorithm Design using PennyLane
Jay Soni, Xanadu

Abstract: Designing quantum algorithms is a very complicated task. There is a whole zoo of algorithmic primitives, and it's often impossible to know which variants of these building blocks, in what combinations, provide the best performance at the cheapest cost. In practice algorithm development is an iterative process. A candidate algorithm needs to be decomposed down to a hardware-compatible gateset to extract its quantum resource cost; then it can be tuned to arrive at the next candidate algorithm before the process is repeated. In this talk, I present some of the experimental tools we have been developing for resource estimation and resource profiling to help accelerate this iterative design process.

Jay Soni completed his BSc in mathematical physics from the University of Waterloo. He's been building quantum software tools alongside the algorithms team at Xanadu for the last 5 years.

Heterogeneous architectures achieve orders-of-magnitude reductions in the space-time resources required for fault-tolerant quantum computing
Yuval Baum, Q-CTRL

Abstract: Over the next decade, quantum hardware is projected to scale to hundreds of thousands of qubits. However, a critical gap remains in architecture design as bottom-up physical constraints—such as the cryogenic and wiring limits—remain largely disconnected from top-down quantum error correction (QEC) code design. We bridge this divide by proposing a modular and heterogeneous quantum computing architecture that integrates task-specific hardware selection with tailored QEC encodings. Our approach provides an explicit microarchitecture for fault-tolerant interfaces between QPUs and quantum memories, bypassing the physical bottlenecks of monolithic quantum computing scaling.
To support this architecture, we present an end-to-end resource estimation framework powered by a custom cross-subsystem compiler. Operating at the scale of 1,000 logical qubits, our tool schedules and orchestrates diverse algorithms down to hardware-specific circuit instructions. Detailed resource accounting reveals orders-of-magnitude reductions in both logical error rates and physical-qubit overhead compared to a monolithic surface-code baseline. As a benchmark, we evaluate the factorization of 2048-bit RSA integers: mapped onto an experimentally demonstrated superconducting grid-coupling topology, our architecture requires 439k physical qubits and 4.9 days. Incorporating modular long-range coupling via qLDPC-based quantum memory further lowers the hardware footprint to just 190k physical qubits with execution time under 10 days, significantly outperforming the space-time factors of existing state-of-the-art baselines.

Dr. Yuval Baum, VP of Quantum Computing Research, has led the research division at Q-CTRL for the past six years. He possesses deep expertise across the entire quantum computing stack, including gate design, circuit synthesis and compilation, error suppression and mitigation, quantum error correction (QEC), and algorithm design. Yuval has a proven track record of collaborating with cross-functional scientific and engineering teams to deliver high-impact quantum technologies and transform cutting-edge research into commercially viable software products. He holds a PhD in Quantum Physics from the Weizmann Institute of Science and completed his postdoctoral fellowship at Caltech’s Institute for Quantum Information and Matter (IQIM).

Compiling the 2D Fermi-Hubbard Ground-State Energy Estimation Algorithm for Active Volume Quantum Architectures
Harriet Apel, PsiQuantum

Abstract: Circuit compilation choices will increasingly depend on details of the underlying architecture rather than solely on generic proxies such as non-Clifford count. We present an active-volume-aware compilation of the ground-state energy estimation algorithm for the two-dimensional Fermi–Hubbard model using quantum phase estimation and Trotterized time evolution. The proposed compilation reduces active volume by up to 3.9×, while the active volume architecture and recent execution-scheduling advances provide additional reductions in estimated resource requirements. Large-scale active volume estimates are obtained using the open-access PsiQuantum Workbench. This provides a practical route to evaluating more detailed resource metrics where analytical expressions become difficult to derive for complex circuits. Together, these results demonstrate the growing importance of architecture-aware compilation and resource estimation for practical early fault-tolerant quantum computing.

Dr. Harriet Apel is a quantum algorithms researcher on PsiQuantum’s Applications and Development team in San Francisco, where she has worked for the past two years. She completed her PhD at UCL with Prof. Toby Cubitt, focusing on the theory of Hamiltonian simulation. This is her first time attending IEEE, where she is presenting work carried out in collaboration with the wider PsiQuantum team.

From resource estimation to actual execution: what if a large-scale fault-tolerant quantum computer hardware magically appeared?
Élie Gouzien, Alice & Bob

Abstract: Resource estimation evaluates a quantum algorithm and a quantum computer architecture at once, and is therefore a powerful instrument for making informed design decisions about both. Its chief virtue is the discipline it imposes: carrying an estimate to the end forces one to specify details that are otherwise easy to overlook, and exposes blind spots in the algorithm and in the architecture alike.

Among these is the logical error model, which cannot be written down without committing to a physical gate set, a connectivity, a physical noise model, and a decoder. One must also know how to apply logical gates in practice, rather than just proving universality of a minimalist gate set. This in turn requires a magic state preparation protocol, specified down to its layout and error model. In architectures that combine several codes to trade encoding rate against access to routing qubits, the division of labor between codes must be settled, together with whether that division is hard-coded or decided on the fly. The same applies to distributed computation, where the availability of buses between subsystems has to be validated before any estimate becomes meaningful. Classical resources, decoding in particular, must likewise be sized correctly.

The more precise the estimate, the closer it comes to a genuine compilation of the algorithm onto a concrete architecture, with the scheduling of fault-tolerant gadgets in space and time as the pivotal step. Refining resource estimates therefore leads to a question that will eventually have to be answered: what exactly is a fault-tolerant quantum computer? Answering it will also bring us closer to running a large-scale fault-tolerant computation.

This talk addresses several of these topics: magic state factories design, ab initio optimal circuit synthesis through translation to a SAT problem, and the estimation of very low logical error rates. It closes with a perspective on the steps needed to bridge the gap between a logical resource estimate and an actual execution.

Dr. Élie Gouzien is a quantum physicist at Alice & Bob, where he works on quantum algorithms, compilation, error correction and resource estimation. He received his PhD from Université Côte d'Azur in 2019 for work on multimode quantum optics, and then joined Nicolas Sangouard's group at CEA Saclay. His recent work target the concrete cost of running cryptographically relevant algorithms on realistic hardware, including estimates for factoring 2048-bit RSA integers with a multimode quantum memory (Phys. Rev. Lett. 127, 140503) and computing 256-bit elliptic curve discrete logarithms with cat qubits (Phys. Rev. Lett. 131, 040602). He also works on other topics around circuit synthesis, quantum error correction and fault-tolerant quantum compilation.

Quantum Computing for Corrosion Simulation and Prevention: Workflows and Resource Analysis
Nam Nguyen, Visa

Abstract: Quantum computing offers a possible route to high-fidelity simulation of corrosion processes that are difficult to treat with classical electronic-structure methods, especially when strongly correlated chemistry appears at metal-environment interfaces. In this talk, I will present a hybrid classical-quantum workflow for corrosion modeling and corrosion-resistant materials design, with two application classes: hydrogen-evolution-driven aqueous corrosion in magnesium and Mg-rich alloys, and oxygen diffusion in high-temperature niobium-rich alloys. Across both case studies, the dominant quantum bottleneck is high-accuracy ground-state energy estimation, which we formulate using qubitized quantum phase estimation in a first-quantized plane-wave representation. I will discuss how these corrosion problems are mapped into quantum subroutines, how the surrounding classical workflow is organized, and what fault-tolerant resources are required for representative models. These results give a concrete picture of the scale of quantum hardware needed for industrially relevant corrosion problems and clarify where quantum computing may fit into practical materials-screening workflows.

Nam Nguyen is a quantum computing researcher at Visa. He previously worked at Boeing Research & Technology, where his work focused on quantum algorithms, resource estimation, and quantum computational workflows for aerospace and scientific applications.

Grand Challenges in Quantum Resource Estimation
Kate Smith, Northwestern University

Abstract: As quantum computing shifts from the NISQ era into early fault tolerance, the central bottleneck is no longer raw hardware capability but how efficiently scarce logical qubits, time, and classical control can be allocated and accounted for. This talk surveys the grand challenges facing quantum resource estimation as the field enters this new regime. Current practice largely relies on manual, workload-specific estimates that struggle to keep pace with heterogeneous hardware, emerging space-efficient QEC codes, and increasingly complex compilation pipelines. To keep up with new developments in fault tolerant computing, quantum resource estimation must evolve into a compiler-driven discipline, similar to how EDA transformed classical chip design, capable of automatically connecting algorithm-level descriptions to concrete costs such as logical error rates, space-time overhead, decoding latency, and classical control burden. How automated, end-to-end resource estimation can serve as a coordinating mechanism across algorithms, error correction, software, and architecture research will be discussed.

Dr. Kate Smith is an Assistant Professor of Computer Science at Northwestern University. She joined Northwestern from Infleqtion, where she acted as a quantum software manager directing projects related to optimized compilation, error mitigation, and simulation of quantum programs on a wide range of quantum technology platforms. Prior to Infleqtion, Kate was an IBM and Chicago Quantum Exchange (CQE) postdoctoral scholar at the University of Chicago Department of Computer Science. She earned her Ph.D. in Electrical Engineering from Southern Methodist University in December 2019.

ECDSA.Fail: Open Autoresearch for Optimizing Elliptic-Curve Point Addition in Shor's Algorithm
Jieyi Long, Theta Labs

Abstract: ECDSA.Fail is an open challenge to optimize secp256k1 point addition, a key arithmetic bottleneck in Shor’s algorithm for attacking elliptic-curve cryptography. Participants ranged from quantum researchers and mathematicians to blockchain developers and software engineers, many of whom had limited prior experience in quantum computing. They combined human expertise with AI coding agents to propose, implement, and evaluate circuit transformations using a shared repository, public leaderboard, and machine-checkable verifier. We call this collaborative research model Open Autoresearch. A community of humans and AI agents searches for improvements against a shared, verifiable objective while openly sharing both successful and unsuccessful results.

Over roughly two months, the community reduced the benchmark from 2,715 logical qubits and 3.96 million average executed Toffoli gates to 1,152 qubits and 1.31 million, a 7-fold reduction in the qubit-Toffoli product that places the result below that of Google's non-public circuit, disclosed only via zero-knowledge proof in March 2026. The community also found a low-qubit solution using only 825 qubits, at the cost of approximately 489 million Toffoli gates. Optimizations identified by the AI Agents included GCD transcript compression, measurement-based uncomputation, operation fusion, register-lifetime reduction, and many local circuit refinements. This talk presents both the resulting circuit improvements and the broader Open Autoresearch methodology that enabled them.

Dr. Jieyi Long is the co-founder and CTO of Theta Labs, where he leads research and development in decentralized computing, blockchain systems, and applied cryptography. Prior to founding Theta Labs, he was a staff engineer at Synopsys, contributing to advanced electronic design automation technologies for large-scale integrated circuits. His current research interests include distributed systems and cryptographic protocols, and his recent work has appeared in leading conferences/journals such as IACR PKC, IEEE DSN, and IEEE TPDS. He received his B.S. in Microelectronics from Peking University and his Ph.D. in Computer Engineering from Northwestern University.