Co-designed transversal STAR architecture, published in PRX Quantum, delivers up to 250× faster execution and roughly 2× fewer physical qubits than conventional fault-tolerant approaches for structured quantum simulation — bringing the megaquop era within reach significantly sooner.
BOSTON, MA — June 1, 2026 — QuEra Computing today announced the publication in PRX Quantum of a new co-designed quantum computing architecture, developed in collaboration with Los Alamos National Laboratory, that significantly reduces the physical resources required for early fault-tolerant quantum simulation. The architecture — called transversal STAR (Space-Time Efficient Analog Rotation) — is co-designed with neutral-atom hardware and is targeted at structured quantum simulation problems in materials science, condensed matter and non-equilibrium dynamics.
The work addresses a central stepping stone on the path to useful quantum computing: the so-called "megaquop" regime, in which an error-corrected quantum computer carries out approximately one million reliable logical operations and begins to perform calculations that are genuinely out of reach for classical methods. Reaching this regime in practice has been a matter of reducing the substantial overhead that quantum error correction imposes on logical computation. Transversal STAR removes two of the most expensive overhead factors at once, extending the original fixed-connectivity STAR proposals [PRXQ 5, 010337 (2024); PRX 15, 021057 (2025)] to reconfigurable architectures.
In standard fault-tolerant architectures, non-Clifford operations rely on resource states known as "magic states" that must be prepared, cultivated, and consumed throughout a circuit. In 2024, QuEra was the first to experimentally demonstrate logical (protected) magic state distillation. For quantum simulation specifically, a second resource requirement is layered on top: the small-angle rotations native to Hamiltonian evolution must be synthesized from a discrete gate set, multiplying the cost. Combined, these two factors make conventional fault-tolerant quantum simulation prohibitively expensive in the early fault-tolerant regime.
Transversal STAR sidesteps both. The architecture prepares small-angle magic states directly, via post-selection-based transversal injection, eliminating the synthesis step. It performs the surrounding Clifford operations transversally — taking advantage of the reconfigurable connectivity and large-scale parallelism that neutral-atom hardware provides — to remove the lattice-surgery routing overhead that limited earlier proposals. The result is an architecture in which the speed of analog rotations and the speed of Clifford gates are matched, eliminating the bottleneck that conventional approaches face.
Through detailed circuit-level simulations with hardware-inspired noise models, the authors show that the surface-code-based version of transversal STAR can simulate local Hamiltonians with a total simulation volume exceeding 600 — defined as the product of the number of logical qubits and Hamiltonian evolution timescales — using approximately 10,000 physical qubits at a physical two-qubit gate error rate of 10⁻³. This represents a 20–40× reduction in overall space-time cost (qubits × clock cycles) relative to the previous best-in-class STAR architecture, which required approximately 20,000 physical qubits and relied on fixed qubit connectivity with lattice-surgery Clifford operations.
The paper also extends transversal STAR to high-rate quantum low-density parity-check (qLDPC) codes, reducing the physical qubit count by a further ~5× — to roughly 1,500-3,000 physical qubits — while preserving most of the time advantage.
"Megaquop-scale quantum simulation is a critical intermediate step towards realizing the full potential of fault-tolerant quantum computation, and reaching it sooner is a matter of architecture as much as hardware," said Sheng-Tao Wang, VP of Algorithms and Applications at QuEra Computing and a corresponding author of the paper. "Transversal STAR shows that when you co-design the algorithm, the code, and the hardware around the structure of the application, you can change what's achievable by orders of magnitude.”
"This work shows the value of pairing computational science expertise with quantum hardware capabilities. By co-designing the architecture around what neutral-atom systems do best — reconfigurable connectivity and large-scale parallelism — we're able to give the simulation community a credible early-fault-tolerant path to longer timescale problems that arise in many-body physics than is currently possible," said Andrew Sornborger, Computer and Computational Sciences Division at Los Alamos National Laboratory and a corresponding author of the paper.
By rebalancing the cost of fault-tolerant quantum simulation, transversal STAR gives researchers, hardware engineers, and quantum software developers a near-term path to:
- Beyond-classical early fault-tolerant simulation: structured local-Hamiltonian simulations across materials science, condensed matter, and non-equilibrium dynamics, at simulation volumes well beyond what NISQ devices and likely what classical HPC can support.
- Reduced hardware footprint: roughly ~2x fewer physical qubits than fully fault-tolerant alternatives for the same simulation workload — and a further ~5× reduction with the qLDPC version.
- Faster execution: approximately 250× faster execution than fixed-connectivity fully fault-tolerant architectures, and ~10× faster than the original fixed-connectivity STAR proposal.
- A clearer co-design path: an example of how algorithmic, code-level, and hardware advances reinforce one another to compress the timeline to beyond-classical scientifically impactful quantum simulation.
The transversal STAR architecture is partially fault-tolerant by design and is targeted at structured quantum simulation rather than arbitrary universal computation. QuEra continues to focus all R&D efforts on delivering large-scale, universally fault-tolerant quantum systems; transversal STAR is a near-term application path that builds on the same neutral-atom capabilities: reconfigurable connectivity, parallelism, transversal gates, that underpin that long-term roadmap.
The publication extends a landmark year for QuEra's fault-tolerant program. In 2025, four Nature papers produced in collaboration with, and in several cases led by, QuEra's academic partners at Harvard and MIT, demonstrated continuous operation of multi-thousand-atom arrays, integrated fault-tolerant architectures with up to 96 logical qubits, the first logical-level magic state distillation, and transversal fault tolerance that reduced runtime overhead by 10–100× for reconfigurable architectures such as neutral atoms.
The paper, "Transversal architecture for megaquop-scale quantum simulation with neutral atoms," is available in PRX Quantum and on arXiv at arXiv:2509.18294. The authors are Refaat Ismail and I-Chi Chen (co-first authors, ordering by coin flip), Chen Zhao, Ronen Weiss, Fangli Liu, Hengyun Zhou, Sheng-Tao Wang, Andrew Sornborger, and Milan Kornjača, with affiliations at QuEra Computing, Los Alamos National Laboratory, the University of Kentucky, Iowa State University, and Washington University in St. Louis.
QuEra will host a live webinar on the work on July 1st 4:00 PM ET / 1:00 PM PT, featuring members of the research team and a live Q&A. Register here.
