With the right mix of Reliability, Availability and Serviceability (RAS) features to meet the most stringent service level agreements (SLAs), HPE Superdome Flex is the perfect server family for demanding mission critical workloads. But did you know that a “fat node” HPE Superdome Flex is also the ideal platform to deliver faster and differentiated insights for your HPC workloads?
Let’s start with a definition of a fat node HPE Superdome Flex. A fat node is a server with a large amount of compute and memory for complex, data-intensive HPC workloads such as simulations, modeling and analytics. Larger HPC workloads are sometimes difficult to split into smaller jobs and allocated to small nodes, and may not provide the same differentiated or unique insights as when you run them as a single job in a single node. If you have a larger HPC job on a smaller server node and run out of memory, you risk losing all of your data and productivity. There’s also the challenge of meeting demanding SLAs for faster insights from your valuable data, which can translate into competitive advantages, increased revenue and reduced costs and risk. If you have limited IT support, you can also configure HPE Superdome Flex as a standalone single fat node, freeing up resources from managing clusters to allow greater focus on discovery. With this platform you don’t have to break your problem across many nodes, distribute your data structures or re-assemble your data sets. It’s just all there, in one single address space and at high speed. In summary, a HPE Superdome Flex fat node is the ideal HPC platform with mission critical features delivering faster and enhanced time-to-insight, completing more jobs in less time and delivering single system simplicity.
HPE Superdome Flex family leverages a symmetric multiprocessing (SMP) design which makes it ideal for scale-up deployments where more memory and compute in a single fat node brings differentiated value to certain complex, data-intensive HPC workloads. For HPC jobs that are more compute-focused and can be run more efficiently across a scale-out HPC cluster, HPE Apollo and HPE ProLiant solutions are ideal.
Let’s cover some specific HPC use cases where HPE Superdome Flex fat nodes are a good fit.
Universities, research institutions, and manufacturers leverage HPE Superdome Flex to tackle the most demanding jobs:
- In big data analytics, the single-system compute power and shared memory of the platform, along with the system’s capacity to accelerate data ingest, enable customers to unlock deeper insights faster.
- In computer-aided engineering, through the use of advanced simulation technology, HPE Superdome Flex has been instrumental in improving the design of wind turbines, aircraft, and racecars.
- And in genomics, this powerful platform has played a key role in DNA research to help predict, diagnose, and treat disease.
HPE Superdome Flex family includes two models, HPE Superdome Flex 280 and HPE Superdome Flex. HPE Superdome Flex 280 is designed for small and medium environments, starting at 2 sockets and scaling up to 8 sockets as a single system in 2-socket increments, using 3rd generation Intel® Xeon® Scalable processors in a glue-less architecture. HPE Superdome Flex is ideal for medium to large environments, scaling from 4 to 32 sockets in 4-socket increments as a single system, and features the 1st and 2nd generation Intel® Xeon® Scalable processors. For more information on the HPE Superdome Flex family of servers and how HPC customers can benefit please visit hpe.com/superdome.