We're focused on de-risking the path to large-scale fault-tolerant quantum computing, and we are doing the work to get there. This work is platform-agnostic — relevant to anyone working on quantum hardware, QEC theory, or benchmarking tools.

🔬: arxiv.org/abs/2508.08188

🧵 1/? Today #quantinuum announced Guppy, the new programming language my team has been developing, and Selene the new simulator that goes with it. This is the software stack that supports our new Helios #quantum computer, and it successors. #quantumcomputing

www.quantinuum.com/blog/built-f...

What we've been working on for the last few years: a modern programming language (Guppy 🐠), compilation tooling and emulation for the next generation of quantum computing, all open source. guppylang.org

Give it a go and tell us what you think!

www.youtube.com/watch?v=dFki...

4. We apply it to a [2,1,1] code on a 20-qubit device using syndrome extraction circuits that protect against leakage by swapping out ancillas and data qubits. We also use Pauli twirling.Their combined effect is a rate of detecting errors stable in time& independent of logical input as QEC promises.

3. In contrast with physical Pauli channel estimation, our decay model incorporates the intermediate classical outcomes obtained from variable length sequences of repeated rounds of syndrome measurements.

That classical data changes with each shot, so we looked into Bayesian post-processing.

2. The protocol is SPAM robust, and simultaneously determines the conditional logical Pauli channels that arise within a region of consecutive syndrome measurements.

Such regions include mid-circuit measurements and resets.

1. Why care about direct LSD (logical syndrome-dependent) error characterization? Because it lets you
✅️ Perform noise-aware decoding, and
✅️ Validate that practical QEC implementations behave as predicted in theory-land.

scirate.com/arxiv/2508.0...
A nice collab with @bencriger.bsky.social @mattgirling @quantinuum.bsky.social

How can we test if quantum devices satisfy the assumptions required for scalable fault tolerance?

📢Our new paper develops a method to directly estimate syndrome-resolved distributions of logical errors arising in syndrome measurements.We implement on @Quantinuum H1-1 & 👀X,Y errors suppressed👇🔗🧵

If you’re running randomized benchmarking on your quantum computer, stop and read our paper!

arxiv.org/pdf/2502.00154

I've been putting together a database of small quantum codes. If you'd like to contribute your parity check matrices, and any references, i'll add them qecdb.org/codes/ Thanks!

Hi #quantum bluesky 👋 In arxiv.org/abs/2503.20870 with #Quantinuum, Fermioniq, Caltech, EPFL, TUM we show Quantinuum's H2 ion-trap #quantumcomputer can simulate dynamical evolution of a 2D Ising model at late times and sizes pushing beyond classical means(MPS, PEPS, sparse Pauli propagation, NN).