Quantum Virtual Machines
summary
HyperQ builds quantum virtual machines that partition hardware by repeating topology tiles, compiles programs to those smaller virtual back ends, and multiplexes them in time and space on real devices. Implemented on the IBM quantum computing service, HyperQ uses greedy packing plus time packing to run multiple circuits concurrently and claims up to an order-of-magnitude gains: ~9× throughput on small jobs, ~4× on medium, and ~26× latency speedup in a sparse arrival model, with no fidelity loss.
AI-written summary · about
chapters (9)
- 0:00Quantum virtual machines intro
- 2:31Current quantum cloud problems
- 4:19Building QVM challenges
- 6:33HyperQ hypervisor design
- 8:00Quantum VM abstraction
- 10:56User perspective & compilation
- 11:58Scheduling compiled programs
- 14:06HyperQ performance evaluation
- 17:37Conclusion & future work
related talks
17mQVM: Quantum Gate Virtualization Machine
18mIntegrating Quantum Computers in High-Performance Computing Environments
23mMake Quantum Computing Useful with Performance Enhancing Infrastructure Software
11mScience & Industrial Innovation With Quantum-Accelerated Highperformance Computers
47mUnboxing Quantum Computing
47mQuantum Computing
47mQuantum in the Cloud: Running Quantum Workloads on Kubernetes
18mAutomated distribution of quantum circuits via hypergraph partitioning