News

Mitsui & Co. and Mitsubishi Electric Benchmark Approximate and Logical QFT on Quantinuum Helios

2 min read

Share

Japanese industrial conglomerates Mitsui & Co. and Mitsubishi Electric have published joint experimental benchmarks evaluating the Quantum Fourier Transform (QFT) on Quantinuum’s 98-qubit Helios trapped-ion quantum computer. Detailed in a co-authored white paper (“Experimental Evaluation of the Quantum Fourier Transform on a Trapped-Ion Quantum Computer“), the team executed both physical-qubit approximate QFT and Steane-encoded logical QFT circuits.

The experiment evaluates hardware scaling and algorithmic utility across two computational regimes:

  • Physical Qubit Approximate QFT (Up to 98 Qubits): The researchers executed approximate QFT circuits up to the processor’s full 98 physical qubit capacity. By setting the small-angle phase rotation truncation parameter (degs=5), the system preserved a non-zero target-state probability (Ptarget​=0.143 at 98 qubits) while applying leakage-detection (LD) post-selection to mitigate environmental physical errors.
  • Logical QFT with Steane Code (Up to 12 Logical Qubits): Using the 7-qubit Steane error-correcting code ([[7,1,3]]), the team instantiated up to 12 logical qubits across 84 physical qubits. Evaluating error-detection post-selection against active error correction, the team observed that error detection achieved higher logical target-state probabilities (PL​target = 0.934 for 4 logical qubits, 0.774 for 8 logical qubits) at the cost of reduced shot acceptance rates (31% at 8 logical qubits, 8% at 12 logical qubits).
  • Logical T-Gate Implementation Trade-Offs: In a two-logical-qubit QFT test, the team compared non-fault-tolerant direct analog rotations with fault-tolerant code-switching state injection (using 30 physical qubits across quantum Reed-Muller and Steane blocks). Direct analog rotation yielded higher output fidelity under current physical noise levels, highlighting the operational overheads associated with full fault-tolerant gate construction.
                    [ Physical vs. Logical QFT Execution Benchmarks ]
                                            │
     ┌──────────────────────────────────────┴──────────────────────────────────────┐
     ▼                                                                             ▼
  Physical Approximate QFT (degs=5)                              Logical QFT (Steane Code)
  • Executed across 18 to 98 Physical Qubits.                   • Scaled from 2 to 12 Logical Qubits.
  • P_target = 0.976 (18q) ➔ 0.143 (98q).                        • Uses 7 Physical Qubits per Logical Qubit.
  • Uses Leakage-Detection Post-Selection.                      • Evaluates Direct Analog vs. Code-Switching T-Gates.

The joint research team developed custom quantum error correction software using Quantinuum’s Guppy® hybrid programming language alongside pytket® compilation toolchains, evaluating cross-layer design trade-offs between physical gate noise, code distance overhead, and post-selection acceptance rates.

Review the technical white paper via Mitsubishi Electric here, and explore executive insights on the Quantinuum Blog here.

In GQI Portal

The players behind the news

The team that writes QCR tracks every company, deal and technology in the GQI Factory, GQI's verified database of the quantum industry. Next up: every story linked to its players, coming to QCR's paid plans.

  • Players Companies and institutions across the quantum industry, by segment.
  • Scorecards How the players compare on hardware, software, funding and more.