HPE Superdome Flex is the perfect “fat node” in HPC clusters that need faster time to insight. Learn how HPE Performance Cluster Manager manages all HPE HPC systems and why it’s now easier to integrate.
A high-performance computing (HPC) cluster is a group of servers, called nodes, which share data and work together on jobs via a fast interconnect. The cluster can feature a variety of nodes to handle different tasks. And this may include “fat nodes,” which HPCwire defines as servers with a large amount of compute and memory for complex, data-intensive HPC workloads such as simulations, modelling and analytics.
Fat nodes are needed whenever an HPC job is too big or complex to break up and run across multiple smaller servers. Fat nodes are also useful when scientists need to accelerate time to discovery, by running more jobs in less time on a giant server with lightning speed.
However, it can be difficult to add fat nodes to your HPC cluster when there is a need to manage them separately from regular servers. Such management burden takes time away from science and may require specialized skills.
HPE Superdome Flex is a perfect fit for the role of fat node, with its massive scalability, shared memory and performance. And now it’s easier to integrate Superdome Flex within your HPC cluster using HPE Performance Cluster Manager (HPCM) software, which provides a unified solution for managing all HPC systems from HPE.
Let’s look more closely at fat node use cases, how system management fits into the HPC software stack, and what HPCM can do for your clusters.
When do fat nodes add value?
Fat nodes can add value to your HPC cluster when you have workloads that are too big or too complex to run across multiple nodes.
Computer-aided engineering (CAE) workloads, for instance, can be highly complex. Examples include electromagnetic simulation aiming to optimize placement of airplane antennae and GPS devices to minimize interference; FEM1 simulation that aims to improve the structural design of wind turbines; or usig computational fluid dynamics to simulate and improve the aerodynamics of race cars.
Medical and agricultural genomics is another field where fat nodes are useful. There, genome mapping might compare billions of small sequences and terabytes of data with a previously assembled genome until complete. And it is much easier to do this with a single large server than to (1) break up the data sets and their complex relationships, (2) re-design applications to orchestrate a cluster of smaller servers, and (3) reassemble everything at the end.
In these cases, investment in a fat node is likely to save time and accelerate insights significantly.
How complex system management impacts HPC
But is complex HPC system management really such a bottleneck for scientists?
Based on years of talking to HPE customers, we find that managing HPC clusters has always been one of the top three challenges in HPC. And today, HPC systems are bigger and more complex than ever before.
An IDC analysis[i] of the HPC software stack identifies two categories of software that teams need to deal with: the applications that run their workloads, and a long list of system software that includes operating systems, management tools, schedulers, plus network and storage software.
HPC system management is crucial to provisioning, managing, and monitoring clusters. According to IDC, software in this category impacts:
- System setup speed
- Time spent on administration
- Cluster performance optimization
- Compatibility with HPC applications software
Having a heterogeneous, standardized management environment — i.e. a cluster of servers that can all be managed as one platform – is needed to achieve optimal HPC performance and cost efficiency.
Keep your HPC systems at peak performance with HPE Performance Cluster Manager
HPE Performance Cluster Manager (HPCM) is a fully integrated system management solution for all HPC clusters and supercomputers from HPE, which helps you achieve a better return on your hardware investments.
Customers choose to manage their clusters with HPCM when they want to:
- Improve day-to-day HPC management – With fast system setup from bare metal, comprehensive hardware monitoring and management, image management and software updates, and power management.
- Manage every aspect of the cluster – Including GPUs, CPUs, interconnect, software, jobs, power and cooling, with customized monitoring, alerts and management actions that fit your needs.
- Manage HPC anywhere – HPCM supports on-premises and hybrid HPC deployments.
HPCM also helps you reduce the time and resources you spend on system administration, while keeping systems as resilient and efficient as possible.
HPCM now supports HPE Superdome Flex
HPCM now also supports HPE Superdome Flex, so you can easily fit these fat node candidates into your HPC clusters – without having to manage them differently.
So, whether you already use HPCM or not, you can build HPC clusters that deliver accelerated time to insight with Superdome Flex-based fat nodes, and achieve optimal cost-performance with a standardized management environment.
To learn more and get started, click below:
Meet Diana Cortes, HPE Marketing Manager, Data Solutions
Diana has spent the past 24 years working with the technologies that power the world’s most demanding IT environments and is interested in how solutions based on those technologies impact the business. A native from Colombia, Diana holds an MBA from Georgetown University and has held a variety of regional and global roles with HPE in the U.S., the U.K. and Sweden. She is based in Stockholm, Sweden. Connect with Diana on LinkedIn.
Compute Experts
Hewlett Packard Enterprise
twitter.com/hpe_compute
linkedin.com/showcase/hpe-servers-and-systems/
hpe.com/servers
[i] High Performance Computing (HPC) Software Stack, Marketplace Research, IDC, August 2022
ComputeExperts
Our team of Hewlett Packard Enterprise server experts helps you to dive deep into relevant infrastructure topics.