Same cyber desk task on IBM Heron and QuEra Aquila
One shared score matrix (9 enterprise assets × 17 cyber desks).
QPC runs it as a native job on each machine — digital on IBM Fez, analog AHS on Aquila —
with public job IDs. That is the universality claim: the layer travels.
Claim boundary:
Same task, hardware-native encodings — not the same circuit, not an IBM scoreline rematch,
not supremacy. Architecture discriminator = coupled vs separable. Domain priority ranks are provisional.
IBM Heron · ibm_fez
Cyber Command Desk · 156Q
Job
d9eephkjeosc73fimti0
Geometry
9 assets × 17 desks + junctions · SamplerV2
Open joint-structure gap
~0.0021 (coupled vs intracontext control)
Top assets (provisional)
Payments · API gateway · LLM agent
Completed · public IBM job
QuEra Aquila · Amazon Braket AHS
Same cyber desk · 153 atoms
Coupled task
…/da791b26-cdc5-4f0d-a55d-7e5849eabbc9
Separable task
…/fffedbb9-626d-472d-9b04-103f1db2657a
Encoding
Local detuning ← scores · geometric bridges
|Δ ZZ_bridge|
0.039 ≥ 0.02 (architecture PASS)
Top assets (provisional)
OT plant · Payments · LLM agent (2/3 name overlap with IBM)
Architecture PASS · dual-publish ready
Also on Aquila — full register 256 atoms
Separate universality benchmark: four contextures on the full Aquila FOV,
coupled vs separable, same global drive. Not a domain desk — a full-size portability witness.