Three-node quantum network, 98-qubit trapped-ion benchmark

Duke University researchers working with IonQ reported this month that they created a three-node quantum network linking atomic qubits with photons. Two days later, Quantinuum published performance results for a 98-qubit trapped-ion processor.

The Duke-IonQ experiment generated a three-node distributed entangled state among atomic qubits connected by photons and addressed the detection loophole, a condition in which detectors miss some particles in an entangled system. The team reported a bounded GHZ state fidelity between 0.841 and 0.881 for the three-node setup, a fidelity range the researchers said exceeds prior demonstrations using diamond color centers or atomic ensembles.

Entanglement generation was about 0.095 successful events per second. The population data reported in the paper came from 687 successful events out of more than 4.1 billion attempts.

Quantinuum released benchmark data for its 98-qubit trapped-ion processor, providing details on both hardware performance and the software stack needed to run larger programmable trapped-ion machines.

Christopher Gannatti, global head of research at WisdomTree, described Quantinuum’s Helios runtime as a real-time classical control stack that maps user-defined ‘virtual qubits’ to physical ions while a quantum program runs, dynamically routes ions, sorts execution batches and supports mid-circuit measurements and conditional logic. He added: ‘That is a meaningfully different kind of achievement than demonstrating a beautiful two-qubit gate in isolation. It is engineering infrastructure for a programmable machine.’

Both announcements came during a roughly two-month decline in technology and AI-related stocks. The WisdomTree Artificial Intelligence and Innovation Fund (ticker WQTM) lists Quantinuum and IonQ among its top holdings; the ETF held about $279.2 million in assets and launched in October. WQTM tracks the WisdomTree Classiq Quantum Computing Index.

Statements from the researchers and companies noted ongoing limitations, including low entanglement generation rates and the need for additional engineering to raise success rates and throughput. The reports add to other recent technical publications on hardware, networking and control systems for quantum computing.