Particle.news
Download on the App Store
8 ARTICLES
·
4w ago

Quantinuum’s 98‑Ion Helios Pushes Quantum Simulation and Error Correction Forward

Helios reports 99.921% two‑qubit fidelity with on‑the‑fly error correction, enabling a 36‑fermion Fermi–Hubbard run that still awaits independent benchmarking.

Rendering of Helios chip
© Quantinuum
Quantum AI launch
Image of a grey square with lots of wires heading into a gold chip with a complicated arrangement of channels on its surface.

Overview

  • The third‑generation trapped‑ion system uses 98 barium‑ion qubits with all‑to‑all connectivity in a 15 K apparatus accessible via the cloud.
  • Using Helios‑1, researchers executed the largest Fermi–Hubbard simulation so far, modeling 36 fermions and detecting signatures of laser‑induced pairing.
  • Quantinuum cites 99.921% two‑qubit gate performance and demonstrated real‑time error correction using Nvidia GPUs.
  • The architecture targets lower overhead for logical qubits, with the company describing schemes that use two ions per logical qubit.
  • External experts praise the progress but stress that broad quantum advantage is unproven and call for rigorous benchmarks; Quantinuum’s roadmap includes a second Helios in Minnesota, a 192‑qubit Sol system in 2027, and a multi‑thousand‑qubit Apollo targeted for 2029.