Dell releases open-source Omnia software to automate deployment of converged AI, data, and HPC workloads.

Dell Technologies Inc., in partnership with Intel Corp. and Arizona State University, has announced Omnia, an open-source software suite designed to help customers manage the convergence of high-performance computing, artificial intelligence and data analytics workloads more easily. Available now, Omnia is essentially a set of Ansible playbooks, or blueprints of automation tasks that can be performed by software without human intervention. According to Dell, the playbooks can be used to speed up the deployment of converged workloads managed by Kubernetes and Slurm, as well as library frameworks, services and applications. Kubernetes is orchestration software for containers that host the components of modern applications, while Slurm is an open-source, fault-tolerant and scalable cluster management and job scheduling system for large and small Linux clusters.

Omnia works by automatically imprinting a software solution onto each server that has been assigned to the workload in question, reducing time to deployment from days to just minutes for a wide range of converged workloads, such as HPC simulations, neural networks for AI or in-memory graphics processing for data analytics. "As AI with HPC and data analytics converge, storage and networking configurations have remained in silos, making it challenging for IT teams to provide required resources for shifting demands," said Peter Manca, senior vice president of integrated solutions at Dell Technologies. "With Dell's Omnia open-source software, teams can dramatically simplify the management of advanced computing workloads, helping them speed research and innovation."

In related developments, Dell is expanding its HPC on Demand services and Dell EMC PowerEdge server line to support VMware virtualization environments. By adding support for VMware Cloud Foundation, VMware Cloud Director and VMware vRealize Operations, Dell said customers can now adopt a hybrid cloud operating model for resource-intensive HPC workloads. Dell's HPC on Demand service provides customers with cloud-based access to its PowerEdge R server systems on a pay-as-you-go basis, enabling customers to tackle data-intensive workloads including product development simulations and genome sequencing.

Mercury Marine, the marine engine division of Brunswick Corp., has already adopted Dell's HPC on demand infrastructure for computer-aided hydrodynamic simulations to create new propulsion systems. The company reports that using Dell's infrastructure enabled it to slash simulation times from 48 hours to under two hours. "We've got a base level of computational availability on hand that we own, and, now with HPC on demand, we're able to run simulations and workloads that just wouldn't be possible on our existing infrastructure," said Mercury Maritime Engineer Simulation Specialist Arden Anderson. "With Dell Technologies and R Systems, we're able to handle more complex workloads and simulations in far less time."

Dell has also added Nvidia A30 and Nvidia A10 Tensor Core graphics processing units as options for its PowerEdge R750, R750xa and R7525 servers, both available starting today. The Nvidia A30 GPUs are designed to handle data-intensive AI inference tasks and mainstream enterprise compute workloads, while the Nvidia A10 GPUs are targeted at mixed AI and graphics workloads.