Thank you + a small open toolkit & honest feedback from an independent Tuna user

[ingo6]

Hi QuTech / Delft team,

I'm a self-taught metalworker (no lab, no enrollment) who's been using your open Tuna chips via Quantum Inspire. First: thank you — making real quantum hardware freely open let someone like me actually learn by doing.

A few small open tools + some honest findings (mostly reproductions on your hardware):

• Readout mitigation (following your readout_correction.py approach) recovers ~18 pp on a full 17-qubit GHZ chain; a clean 3-qubit block reaches ~97% readout-corrected.
• Redundancy beats length at equal qubit budget (~1.45×) by quarantining the worst qubit.
• 2-of-3 units with disjoint active couplers keep crosstalk from leaking (~2% to the independent-units formula, across 4 geometries on two chips).

One small observation that might be useful: on Tuna-9, q8's T1 fluctuates between ~0.2 and ~7.4 µs over 48 h (consistent with the TLS you flagged switching on/off) — it works surprisingly well as a live "canary" for local noise.

Everything is here (open, Apache-2.0, honest — SIM vs TUNA always labeled): https://github.com/ingo6/tuna-quantum
I only ever read the public dashboards; never touched anything gated.

Thanks again for the open door.
— Ingo (ingo6)