UniCloud™ plugs Univa® Grid Engine™ into any cloud management system or service. The software enables public and private cloud resources to be utilized with existing HPC workflow and seamlessly extends the reach of cloud systems to the largest HPC user community - Grid Engine - with over 10,000 deployments.
Integrating Grid Engine Workflow to Clouds
UniCloud simplifies the configuration and management of HPC compute infrastructure across any cloud - private, public and hybrid - by plugging into leading systems management products and clouds. With UniCloud, organizations will reduce mistakes through the automation and synchronization of management tools and eliminate coding through the integration of existing workflow into cloud resources.
The advantages of meshing grid and cloud systems management include ubiquitous resources, system and workload mobility, and the unleashing of advanced features like seamless cloud bursting and check-pointing.
The unique adapter plug-in design of UniCloud makes connecting new management tools or cloud services simple. Out-of-the-box UniCloud includes adapters to all the leading public cloud and private cloud providers.
Univa brings improved simplicity to HPC in the Cloud. Never before has extending Grid Engine to the cloud been easier or more affordable. UniCloud offers the flexibility of preserved workflow integration, choice of configurations,(CloudClusters, Cloud Bursting, Hybrid Could) choice of cloud provider to provision workload, and the simplest one click HPC offering.
Univa creates workload mobility with Multi-Cloud provisioning. By removing the dependencies on where jobs can run - rather than locking users in to specific hypervisors or cloud APIs - UniCloud can provision Grid Engine to any server or hypervisor, anywhere.
UniCloud optimizes resource utilization by ensuring applications are matched to the optimal data center resources. By connecting to private cloud management systems, Univa helps to accelerate the benefits of cloud while reducing the barriers and risks.
Using its built-in policy engine, UniCloud make decisions about where applications should run in the infrastructure and what machines and memory combinations are needed to support optimal performance. If those machines (physical or virtual) do not exist, UniCloud requests or creates them to meet the needs of the application, ensuring memory and cores are not wasted.
Once new machines are created and the infrastructure has been shaped to meet the needs of the application, UniCloud can then automate the configuration of those new machines so all software is installed and ready to deploy - including the application.
Finally, UniCloud dynamically optimizes this new infrastructure to automatically respond to user and project demands by continually repeating the process described above.