Author: Michael Jackson, President, Adaptive Computing
Convergence: The Second Key to Success in Complex Compute Environments
Welcome back to the “Seven Keys to Success” blog series. In our last post we talked about consolidation, or the process by which physically separate compute environments are brought together. The next step is to determine how those workloads will be able to coexist. This leads us to our next key – Convergence.
Within compute environments, convergence refers to the merging of different classes of workloads from diverse disciplines, such as High Performance Computing (HPC), High Throughput Computing (HTC)/Grid, Data Center, and Big Data. Convergence further becomes a necessity when results from one workload need to be processed by another application. Traditionally, this has required the implementation of complex and time-consuming data processes, as IT personnel were required to integrate across siloed environments.
Moab, however, breaks down those barriers, by allowing different classes of workloads to coexist on a shared platform. For example, sometimes High Throughput Computing (HTC) workloads, which involve many small tasks, require subsequent processing through High Performance Computing (HPC), which handles traditional batch, interactive and large parallel workloads. When Moab is integrated with Nitro, it is capable of managing both HTC and HPC workloads on the same system. It allows you to effectively convert what used to be processes from completely separate siloes, into an integrated data process.
Moab also aids in convergence with traditional Data Centers. Its workflow capabilities enable the importation of data from multiple data services in the Data Center. Moab can then transform the data into HPC/HTC tasks, interact with Data Center databases, and initiate post processing tasks. Moab’s ability to converge these diverse computing environments facilitates the delivery of services by organizations, whose numerous data types and processes are increasingly becoming the norm.