Research

Qoro Quantum opens managed GPU simulation tier for its Divi Python framework

Qoro Quantum has made its cloud-based GPU simulation tier generally available, according to Quantum Computing Report. The managed service runs tensor-network and noisy circuit simulations on remote NVIDIA A10G and T4 GPU clusters, dispatched from local scripts through Qoro's Python framework Divi (qoro-divi). In Qoro's benchmarks of a 64-qubit (simulated, not physical hardware) Ising chain quench, GPUs finished late-time entangled circuits in about 3 minutes versus 11 to 15 minutes on 16-vCPU instances, and at bond dimension 512 all 100 circuits finished on GPU in under 9 minutes each while CPU instances hit a 20-minute limit. A 2,560-circuit noise sweep across 8 two-qubit error rates finished in 43 minutes on parallel cloud CPUs, versus 15.6 hours sequentially; free developer accounts get 1,000 compute credits.

Why it mattersSimulating 64 qubits with a full statevector would need roughly 295 exabytes of RAM at double precision, so developers rely on tensor-network methods that usually require manual cluster setup. Qoro says its service removes that setup, though the benchmark figures are the company's own.

Read the full story at Quantum Computing Report →

Summarized by AI from the linked article.