Windows Azure HPC Scheduler
Windows Azure HPC Scheduler makes it easy to migrate and build compute-intensive parallel applications in Windows Azure. In addition, it enables you to run virtual nodes on-demand in the cloud, thus providing access to computing resources to handle peak or unpredictable workloads. So you can use smaller on-premise clusters and rely on Windows Azure to provide additional scale when required.
There are many practical uses for the Windows Azure HPC Scheduler including:
MPI applications
MPI(Message Passing Interface)is a standardized and portable message-passing system designed by a group of researchers from academia and industry to function on a wide variety of parallel computers. MPI applications are the dominant application type in the traditional high-performance computing market for computational fluid dynamics, finite element analysis, and computational chemistry. By bringing Microsoft’s implementation of the MPI stack to Windows Azure and tuning it for the environment, a wide range of traditional HPC codes can run in Windows Azure.
Cluster SOA applications
Cluster SOAis a programming model for Service Oriented Applications (SOA) in a cluster environment. Cluster SOA applications are the leading application type in the financial services segment of the high-performance computing market. Insurance companies and investment banks use Monte Carlo simulation and similar techniques to price insurance policies and many types of financial instruments. The Cluster SOA runtime enables rapid application development using familiar languages with easy migration to Windows Azure.
Capabilities and Benefits of HPC Scheduler
On demand scalability and resource management
- By being able to run as many nodes as you want in Windows Azure, you have practically unlimited scalability.
- Running applications entirely in Windows Azure provides access to vast computing resources on a pay-per-use basis, eliminating the risk of a large upfront investment in infrastructure for organizations with tough computational challenges.
- Running HPC applications in Windows Azure frees your resources to focus on solving complex problems rather than investing in capital and skills to administer an on-premise cluster.
Cluster Management
- A single ‘pane of glass’ to manage on-premise nodes as well as Windows Azure nodes simplifies monitoring and management.
Pricing and Metering for HPC Scheduler
The Windows Azure HPC Scheduler is provided to customers at no additional cost. To download and get in-depth information on Windows Azure HPC Scheduler, please visit the Windows HPC Azure Scheduler page.