Moab Adaptive HPC Suite™ is a workload management solution powered by a unique multi-dimensional and predictive decision engine that dynamically manages workloads and resources so HPC environments self-optimize to produce more results, with greater speed and efficiency. Moab Adaptive HPC Suite overcomes the challenges of meeting changing workload demands by dynamically adapting the resources to the changing requirements of user workloads, streamlining processes, enforcing SLAs, reducing power consumption, and achieving organizational objectives.

With Moab Adaptive HPC Suite you can overcome key HPC challenges including:

• Delays to workload start and end times slowing results.
• Inconsistent delivery on service guarantees and SLA commitments.
• Under-utilization of resources.
• Need to accommodate workload surges.
• How to efficiently manage workload across hybrid systems of GPGPUs, mixed operating systems, hardware, etc.
• How to simplify job submission & management for users to maximize productivity
• How to reduce power usage and costs to be in line with actual fluctuations in workload and resource usage

A Dynamic Decision Engine to Manage Today’s Complex HPC Environments

Every new technology that accelerates workload execution or requires different run-time environments adds to the complexity of managing state-of-the art HPC facilities. Node counts are growing from thousands to tens of thousands and supporting multiple operating environments and specialized technologies such as GPGPUs makes maintaining operational efficiency increasingly challenging. Even when the capability of existing resources is sufficient, re-provisioning and re-allocating them to meet user requirements is labor intensive and simply takes too long. Efficiently mapping workloads to available resources is no longer enough, resources need to be dynamically adapted to meet user’s workload requirements.

Moab Adaptive HPC Suite not only schedules and allocates the optimal diverse set of resources for massive scale HPC environments, but also determines when the OS mix should be modified for maximum efficiency. Based on defined policies, service-level agreements, current and projected workload, it automatically triggers the OS change using your preferred OS-modification technology, such as dual boot, diskful, or stateless (diskfree) provisioning.

Moab Adaptive HPC Suite enables maximum resource utilization and increased ROI for both new and existing clusters. The mixed-OS cluster model also consolidates administration and centralizes job submission across OS platforms. Administrators can easily manage the policies and workload for multiple OS environments using the unified console. Moab can also make the dual-OS nature of the cluster transparent to end users by applying application and workload information that ensures jobs run on the correct OS without the need for users to specify an OS.

Managing the World’s Top Systems, Ready to Manage Yours

Moab manages the world’s largest, most scale-intensive and complex HPC environments in the world including 40% of the top 10 supercomputing systems, nearly 40% of the top 25 and 36% of the compute cores in the top 100 systems based on rankings from the www.Top500.org. So you know it is battle-tested and ready to efficiently and intelligently manage the complexities of your environment.

Related Solutions:

High Performance Computing, Data Center Management