A pharmaceutical company is designing the Azure compute infrastructure to host a molecular modeling simulation application. The simulations run continuously for up to 72 hours. If a virtual machine is interrupted or terminated during a run, the progress is lost, and the entire simulation must restart from the beginning. The application requires high-performance CPU capabilities and low-latency node-to-node communication. The finance team recommends using Azure Spot virtual machines to minimize compute costs. Which compute virtualization strategy should you recommend to meet the requirements of the simulation application?
- ADeploy the simulation application on H-series Azure Spot virtual machines.
- BDeploy the simulation application as serverless tasks using Azure Functions on a Consumption plan.
- Deploy the simulation application on standard H-series virtual machines.Cevap
- DDeploy the simulation application on D-series Azure Spot virtual machines.
Cevap
Deploy the simulation application on standard H-series virtual machines.
Deploying the simulation application on standard H-series virtual machines is the correct strategy because H-series virtual machines are optimized for High-Performance Computing (HPC) workloads, providing the necessary CPU performance and low-latency InfiniBand networking. Standard VMs do not have an eviction policy, guaranteeing that the 72-hour simulation runs will complete without interruption.
Adım Adım Çözüm
Anahtar Kavram
Selecting Azure VM series and pricing tiers based on performance, network latency, and availability constraints.
Tahmini Süre:1m 30s