Rescale's control plane brings in an intelligence-driven logic and control layer between R&D workloads and the underlying infrastructure, facilitating flexibility, performance, and efficiency, unlike traditional static HPC environments
FREMONT, CA: Rescale, the pioneer in enterprise big compute, has announced the availability of the first hybrid and multi-cloud intelligent control plane to provide high-level performance, architectural control, and R&D productivity.
Enterprises often confront the challenge associated with fast-growing R&D workloads and specialized hardware architectures, all while managing cost and performance for hybrid cloud operations. The Rescale intelligent control plane merges software-defined computing, hardware-based acceleration, and performance intelligence to offer the industry-best high-performance computing solution for each project per business goals.
"We have a variety of specific and intensive compute requirements that Rescale was able to optimize and help us choose which types of processors best fit the workloads that we run," said Sensatek Propulsion Technology CTO, Josh McConkey. "With Rescale, we can dynamically allocate and maximize our HPC resources, ultimately reducing our projected 2,200 simulation hours down to 300, or nine months of work down to two months."
Rescale's control plane brings in an intelligence-driven logic and control layer between R&D workloads and the underlying infrastructure, facilitating flexibility, performance, and efficiency, unlike traditional static HPC environments.
Rescale intelligence includes real-time cloud provider infrastructure pricing worldwide, software licensing schemes, and the Rescale Performance Index, an always-current library of benchmarks on application-specific performance. Enterprise IT can leverage these parameters in defining policies, and track project spending for software and hardware for full-stack economic visibility.
Key Rescale intelligent control plane components are Rescale API, Rescale Intelligence, Operational Workflows, and R&D Workflows. Rescale API estimates job submission and admin controls for unified hybrid and multi-cloud as well as controls resources managed by schedulers, such as SLURM, LSF, PBS Pro, and leading cloud providers like AWS, Azure, GCP, and Oracle Cloud. Rescale intelligence is designed for software licensing intelligence and global infrastructure intelligence on multi-cloud pricing, maturity, capacity, and workload-specific performance. Operational Workflows define security, architecture, performance, and budgeting policies based on the Rescale API and Rescale intelligence. Finally, R&D Workflows are easy-to-use tools for researchers to run jobs and visualizations, collaborate on results, without making infrastructure decisions.