大规模混合量子-经典工作流架构.pdf

1、 2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.C M P 3 0 4Sebastian Stern,PhDTyler Takeshita,PhDArchitectures for Hybrid Quantum-Classical Workfl

2、ows at Scale(he/him)Sr.WW Specialist SA QuantumWWSO Advanced ComputeAWS(he/him)Sr.Quantum Applied ScientistWWSO Advanced ComputeAWS 2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.AgendaIntroduction:Quantum-classical computingArchitecture#1:Fully-managed hybrid quantum-classical c

3、omputing with Amazon Braket Hybrid JobsCustomer use case:Transition from single-service to multi-service workload executionArchitecture#2:Cloud-native compute pipeline for QMC-type workloads with Amazon Braket and AWS BatchArchitecture#3:Integrating AWS PCS and Amazon Braket to run quantum-classical

4、 workloads with SlurmWrap up 2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.Introduction to quantum-classical computing 2025,Amazon Web Services,I

5、nc.or its affiliates.All rights reserved.Accelerated computing on AWSComplexity of workloads is driving a need for higher performanceAmount of compute acceleration needed is driving a need for higher efficiencyVarious workloads are driving a need for multiple compute options 2025,Amazon Web Services

6、,Inc.or its affiliates.All rights reserved.AWS EC2 accelerated compute instancesClassical Compute-GPU,AWS ML Accelerators,and FPGA-based EC2 instancesTrainium acceleratorInferentia acceleratorGB200,B200,H200,H100,A100,L4,L40SA10G,T4Gaudi acceleratorRadeon GPUXilinx acceleratorXilinx FPGADL1F1VT1AI/M