Questions? Feedback? powered by Olark live chat software
Spring over navigation

Konceptuel arkitektur: HPC cluster deployed in the cloud

Customers with HPC applications that can scale to thousands of compute cores, extending on-premises big compute or running as a 100% cloud native solution. This entire HPC solution, including the head node, compute nodes and storage nodes runs in Azure, leaving you with no hardware infrastructure to maintain.

Løsningen er bygget på følgende styrede Azure-tjenester: VM-skaleringssæt, Virtuelt netværk, Lager. Disse tjenester kører i et miljø med høj tilgængelighed, der rettes og understøttes, og det giver dig mulighed for at fokusere på din løsning i stedet for på det miljø, de køres i.

HPC cluster deployed in the cloud OR Availability Set ARM template Script file ClusterHead Node Virtual Machines VM Scale Set RDMA Network A8, A9, and H SeriesVirtual Machines Storage Virtual Network

Implementeringsvejledning

Produkter Dokumentation

HPC Head Node

The HPC head node runs on-premises or within Azure as a Virtual Machine.
Azure virtual machines of size A8 / A9 provide the HPC compute nodes, running Windows or Linux operating system.
RDMA networking available with A8 and A9 instances is utilized to achieve high bandwidth and microsecond latencies between compute nodes.
Storage nodes can also be run within Virtual Machines.

VM-skaleringssæt

Availability Sets ensure that the application is available and resilient to updates and hardware faults.
Virtual Machines communicating via RDMA are placed within the same Availability Set.

Virtuelt netværk

Virtual Networks provides IP connectivity between the head node, compute nodes and storage nodes.

Lager

Azure Storage blobs store the disks backing the Virtual Machines and provides long-term storage of unstructured data and executable files used by the HPC application.

ARM templates

ARM templates or script files are used to deploy your application to the HPC environment.

Relaterede løsningsarkitekturer

On premises HPC implementation bursting to Azure

Customers with HPC applications that can scale to thousands of compute cores, extending on-premises big compute or running as a 100% cloud native solution. This HPC solution can extend its computational capacity by leveraging the compute intensive instances of Virtual Machines running in Azure and accessed via Express Route or VPN.

Få mere at vide
Big Compute-løsninger som en tjeneste

Kunder med HPC-programmer, der kan skalere til tusindvis af beregningskerner, udvider Big Compute i det lokale miljø eller kører som en 100 % oprindelig cloudløsning. Denne HPC-løsning bliver implementeret i Azure Batch, der gør det muligt at jobplanlægge, autoskalere beregningsressourcer og administrere udførsel som en platformtjeneste (PaaS), der reducerer kodning og vedligeholdelse af HPC-infrastrukturen.

Få mere at vide