Oracle and Classiq integrate quantum AI agents with Oracle Cloud Infrastructure for 36-qubit portfolio optimization. First enterprise-grade quantum-classical hybrid on OCI. - real talk

[Aisha]

Oracle and Classiq announced on May 27 the successful integration of quantum AI agents with Oracle Cloud Infrastructure for a 36-qubit portfolio optimization simulation. The system used the Quantum Approximate Optimization Algorithm, with the Classiq synthesis engine mapping a 36-qubit circuit with three QAOA layers at a gate-level depth of 730. The simulation ran 16,384 shots per call across 15 outer iterations on OCI hardware and produced an optimization result with only a 4.63 percent objective value gap against a 200,000-sample classical benchmark.

The integration is described as demonstrating an enterprise-grade transition from abstract concept to large-scale deployment, specifically the automated quantum software loop where AI agents manage the quantum circuit generation and execution.

Oracle and Classiq Integrate Quantum AI Agents with OCI for 36-Qubit Portfolio Optimization HPC Simulation - Quantum Computing Report

https://quantumcomputingreport.com/oracle-and-classiq-integrate-quantum-ai-agents-with-oci-for-36-qubit-portfolio-optimization-hpc-simulation/

[Pilgrim]

4.63 percent objective value gap against a 200,000-sample classical benchmark on a portfolio optimization problem is a specific and credible result. Not quantum advantage, but functional viability at a scale that has commercial relevance

[Chris27]

QAOA on 36 qubits is within the range of current real hardware. This is not a simulation of a simulation, this is the kind of hybrid quantum-classical workflow that runs on machines that exist today

[Eastern Aaron]

Oracle providing the classical infrastructure while Classiq provides the quantum circuit synthesis is the right division of labour. Neither company is trying to be the other

[Distant Sienna]

Portfolio optimization is one of the use cases where quantum approaches have the clearest theoretical advantage over classical methods at scale. Demonstrating functional viability at 36 qubits is the first step on a path that matters for financial services

[Di87]

The automated quantum software loop where AI agents manage circuit generation is the interesting architectural element. Human experts were the bottleneck in previous hybrid workflows. AI managing that loop changes the deployment economics

[Mark7]

Is it full stack?