This week Microsoft and IBM have tied up an agreement that will provide an enhanced level of testing and support for both hybrid and pure IaaS deployments of Spectrum Symphony and Spectrum LSF into Microsoft's Azure public cloud. IBM Spectrum Symphony and Spectrum LSF provide enterprise-class workload management for distributed high performance computing and analytics and have an established brand within their respective markets.
The combination of these market leading solutions will provide our customers with a greater level of agility and scalability required to meet the demands of the industry. This is most apparent in financial services where the rapid developments in regulation have not only impacted the profitability of banks but also put huge strains on their aging infrastructure. As budgets get cut and risk modeling requirements increase, presenting such a partnership between IBM and Microsoft will enable our banking customers to burst from on-premises into Azure.
In the test chart results below, there are several different test runs being made with a varying number of tasks submitted to the available 9920 cores.
You can see that Symphony scales well and consistently over a wide range of tasks being run, from 100 all the way up to nearly one million tasks. To execute 992,000 ten second tasks took approximately 1000 seconds. The chart shows that Symphony running on a 9920 core Azure system can execute 3.4 million ten second tasks per hour.
For customers who bring their own Spectrum Symphony or Spectrum LSF licenses to Microsoft Azure, IBM will continue to deliver support directly just as it does when those licenses are deployed on customers’ premises. To request technical support for use of IBM software on Azure in BYOL/BYOSL scenarios, contact IBM technical support or refer to the IBM Support Handbook.
Join Microsoft and IBM in Salt Lake City on Wednesday 16th at 2:45pm to learn how combining these market leading solutions will provide our customers with the performance and scalability needed for their most demanding workloads. In this session we will review architecture, best practices, and a review of the scalability and performance testing results.
Quelle: Azure
Published by