## Does QuTech's Tuna-17 Make Europe's Best Publicly Accessible Superconducting Qubit?

QuTech's Tuna-17 — a 17-qubit, 24-tunable-coupler superconducting processor built at TU Delft — went live on the Quantum Inspire cloud platform on July 10, 2026, with free, uncapped access for any researcher, student, or educator worldwide. It is the third processor in the Tuna series released within a single year, following Tuna-5 and Tuna-9, and the team has already confirmed the next target: Tuna-28.

The headline numbers: 17 physical qubits, 24 tunable couplers, support for mid-circuit measurements, and a platform that permits more than 100,000 shots per job — a configuration deliberately engineered to reduce statistical uncertainty in [quantum error correction](https://quantumintel.tech/glossary/fault-tolerant-quantum-computing) experiments. The system supports Qiskit and PennyLane SDKs and runs automatic self-calibration with continuous public performance reporting.

The supply chain behind the machine is entirely European: QuTech/TU Delft handles QPU design, fabrication, cloud integration, and system maintenance; TNO provides the compilation layer and co-designs the QPU; Orange Quantum Systems delivers the Quantum OS and calibration software; Qblox supplies control electronics; Delft Circuits provides cryogenic cabling; and QuantWare supplies cryogenic packaging and amplifiers. This is not a single-vendor system dressed up as open architecture — it is a genuinely distributed build across six named organizations.

---

## What Tuna-17 Is Actually Built to Do

The 24 tunable couplers are the key architectural feature worth dwelling on. Tunable couplers allow two-qubit gate parameters to be adjusted dynamically, which is essential for suppressing residual ZZ coupling — a dominant error source in fixed-frequency transmon architectures. More couplers than qubits (24 vs. 17) signals a design philosophy prioritizing gate fidelity and QEC compatibility over raw qubit count.

Mid-circuit measurement support — the ability to measure a subset of qubits during a circuit and feed results forward — is a prerequisite for most practical [quantum error correction](https://quantumintel.tech/glossary/fault-tolerant-quantum-computing) protocols. Without it, you cannot implement syndrome extraction in a surface code. QuTech explicitly frames Tuna-17 as enabling "many quantum error correction experiments," which is a carefully worded claim: the system is a research platform for QEC, not a demonstration of [fault-tolerant quantum computing](https://quantumintel.tech/glossary/fault-tolerant-quantum-computing) itself. That distinction matters for enterprise buyers evaluating European hardware.

The universal gate set also supports [NISQ](https://quantumintel.tech/glossary/nisq)-era algorithms, including what the release describes as "low-depth factorisation" experiments. No specific gate fidelity, T1/T2 coherence times, or quantum volume figures appear in the source material — a notable omission that prospective users should probe directly through the Quantum Inspire performance dashboard before designing experiments.

---

## The Open-Architecture Thesis: Strategic, Not Just Philosophical

QuTech's framing of Tuna-17 as an "open-architecture" system is doing real industrial work here. The dominant model in commercial quantum computing — vertically integrated stacks where the QPU vendor controls hardware, firmware, compilers, and cloud access — creates lock-in and limits third-party component suppliers. QuTech is explicitly betting against that model.

The announcement of a dedicated open-architecture system integration unit being established in Delft, in collaboration with European industrial partners, signals that QuTech intends to productize this approach rather than keep it academic. This is structurally significant for the European quantum supply chain: if open-architecture procurement norms take hold among European publicly funded quantum programs, it creates a viable route to market for component specialists like Qblox and QuantWare that would otherwise be squeezed out by vertically integrated incumbents.

For context, [IBM Quantum](https://quantumintel.tech/companies/ibm) and [Google Quantum AI](https://quantumintel.tech/companies/google-quantum-ai) both operate closed stacks where the QPU, control systems, and cloud layer are proprietary. QuTech's bet is that interoperability will matter more as systems scale — a reasonable hypothesis, though one that has yet to be validated at commercially relevant qubit counts.

---

## Iteration Velocity: The Tuna Roadmap Signal

Three processors in one year — Tuna-5, Tuna-9, Tuna-17 — with Tuna-28 already in development is the most underreported data point in this announcement. Iteration speed in superconducting hardware is constrained by fabrication turnaround, cryogenic bring-up time, and calibration cycles. A roughly doubling of qubit count per release, sustained across three generations, suggests the DiCarlo Lab has achieved meaningful process discipline, not just one-off engineering success.

Tuna-28, if it maintains the architectural approach of Tuna-17 (tunable couplers, mid-circuit measurement, universal gate set), will push into territory where small surface code patches — logical qubits encoding one [logical qubit](https://quantumintel.tech/glossary/logical-qubit) across multiple physical qubits — become experimentally accessible. That is the inflection point where European academic QEC research transitions from theory-adjacent to hardware-validated.

The Tuna-17 development was supported by the HectoQubit consortium under QDNL Phase 2 CAT-1 funding and is a milestone within the EU Flagship OpenSuperQPlus Delft demonstrator. These funding designations matter for timeline expectations: EU Flagship programs operate on multi-year roadmaps with defined deliverables, which means Tuna-28 has institutional momentum behind it, not just research ambition.

---

## Access and Workforce Development

The platform permits more than 100,000 shots per job with no usage caps — a configuration that directly addresses one of the most common friction points in academic quantum hardware access, where shot limits force researchers to choose between statistical rigor and job queue time. Quantum Inspire backends are already integrated into TU Delft's Quantum Information Science and Technology Master's programme curriculum.

Free, uncapped access to a QEC-capable processor is a meaningful workforce development instrument. The bottleneck in quantum error correction research is not theory — it is hands-on experience with hardware that actually supports the protocols. Tuna-17 addresses that gap at zero marginal cost to the user.

---

## Key Takeaways

- **Tuna-17 is a 17-qubit, 24-tunable-coupler superconducting processor** developed by the DiCarlo Lab at QuTech/TU Delft, now live on Quantum Inspire with free, uncapped access.
- **The 24 tunable couplers exceed the qubit count**, indicating a design priority on gate fidelity and QEC experiment compatibility over raw qubit numbers.
- **Mid-circuit measurement support** makes Tuna-17 capable of running syndrome extraction experiments — a prerequisite for surface code QEC work.
- **Three Tuna systems shipped in one year** (Tuna-5, Tuna-9, Tuna-17), with Tuna-28 already in development, demonstrating a sustained iteration cadence.
- **Six European organizations** contributed to the hardware stack: QuTech/TU Delft, TNO, Orange Quantum Systems, Qblox, Delft Circuits, and QuantWare.
- **No gate fidelity, T1/T2, or quantum volume figures** were published in this announcement — a gap users should verify via the Quantum Inspire performance dashboard.
- **A dedicated open-architecture integration unit** is being established in Delft, signaling commercial intent beyond academic demonstration.

---

## Frequently Asked Questions

**What is Tuna-17 and who built it?**
Tuna-17 is a 17-qubit superconducting quantum computer with 24 tunable couplers, developed by the DiCarlo Lab at QuTech/TU Delft. It was built collaboratively with TNO, Orange Quantum Systems, Qblox, Delft Circuits, and QuantWare — all European organizations — and is accessible for free through the Quantum Inspire cloud platform.

**How does Tuna-17 compare to IBM or Google superconducting systems?**
The source material does not provide gate fidelity or coherence time data, making direct benchmarking impossible from this announcement alone. Structurally, Tuna-17's open-architecture design and European supply chain differentiate it from IBM Quantum and Google Quantum AI's vertically integrated stacks. Its 17-qubit count is modest relative to current commercial leaders, but its QEC-oriented design (tunable couplers, mid-circuit measurement) makes the comparison to raw qubit counts somewhat misleading.

**What quantum error correction experiments can Tuna-17 support?**
Tuna-17 supports mid-circuit measurements and a universal gate set, enabling syndrome extraction circuits used in small surface code and stabilizer code experiments. QuTech describes it as enabling "many quantum error correction experiments" — it is a research platform for QEC work, not a demonstration of fault-tolerant operation.

**How do I access Tuna-17?**
Through the Quantum Inspire platform (quantuminspire.com), with no usage fees and no shot caps. The system supports Qiskit and PennyLane, so existing quantum software workflows can be adapted without significant overhead.

**What comes after Tuna-17?**
QuTech has confirmed Tuna-28 is already in development. If the iteration cadence of the past year continues, Tuna-28 could reach a qubit count where small logical qubit demonstrations using physical qubit arrays become feasible.