The HPE GreenLake platform powers your projects with a cloud experience that lets you avoid the costs, risks, and latency of moving your data to and from the public cloud, while meeting the compliance and regulatory requirements for privacy and data sovereignty. Join us at GTC to learn more.
In addition, for many organizations, high performance computing can be costly, as well as complex and time-consuming to manage. Public cloud is typically not the best option, because workloads depend on low-latency data movement, powerful processing, and high availability of server resources. Luckily, for IT and OT teams supporting AI/ML and HPC workloads, as well as for teams managing the data and ML model lifecycles, there is an option with the flexibility, pay per use, and scalability of public cloud on premises.
The HPE GreenLake edge-to-cloud platform is a fully managed on-premises solution with a cloud experience that lets you avoid the costs, risks, and latency of moving your data to and from the public cloud, while meeting the compliance and regulatory requirements for privacy and data sovereignty.
HPE GreenLake for MLOps brings DevOps agility to every stage of the ML lifecycle, and frees up your data scientists to concentrate on data science by allowing them to easily access open-source or third-party independent software vendor (ISV) tools, as well as required data sources across their organization. The new release of the solution runs a variety of NVIDIA accelerators, including the A100 Tensor Core GPU for AI/ML training and inferencing use cases in both HPE Apollo and HPE ProLiant servers, which are all NVIDIA-Certified.
NVIDIA NGC catalog is a huge software ecosystem that contains several production-tested AI container libraries and continues to expand at a rapid rate. HPE and NVIDIA worked closely to integrate two of these libraries, for deploying at scale, that are tested and made available in the HPE GreenLake marketplace: NVIDIA RAPIDS Accelerator for Apache Spark and the NVIDIA Triton Inference Server. These containerized workloads deliver faster processing and response, faster model training and, ultimately, faster insights (when compared to CPUs). Part of HPE GreenLake MLOps, HPE Ezmeral software simplifies the development, scaling, and management of Kubernetes clusters across hybrid environments.
HPE GreenLake for High Performance Computing now includes support for NVIDIA NVLink and NVIDIA A100, A40, A30 Tensor Core GPUs in increments of 2-4-8 accelerators, ready to quote, price, and deliver. With HPE GreenLake, customers can choose an on-premises HPC solution running on NVIDIA GPUs with all the flexibility, scalability, and utility-like consumption of the cloud. Customers can increase their agility with pay-per-use pricing and a preinstalled buffer capacity that is ready to provision when needs grow. Whether demand for HPC resources spikes suddenly or grows steadily, customers are always ready to meet their new needs.
Customers can design their own HPC infrastructure solution within HPE GreenLake using industry-leading HPE technologies. Alternatively, they can standardize their service with pre-sized configurations that offer self-service and are managed for them. With access to the latest technology available in the market, HPE can also buy out—and recycle—customers’ existing infrastructure, helping them meet sustainability targets.
The newest member of the NVIDIA GPU portfolio is the NVIDIA H100 Tensor Core GPU, powered by the NVIDIA Hopper architecture, which delivers the next massive leap in NVIDIA’s accelerated compute data center platform, securely accelerating diverse workloads in every data center, from small enterprise workloads to exascale HPC and trillion-parameter AI. The NVIDIA H100 delivers unprecedented performance, scalability, and security to every data center, accelerating everything from Exascale and trillion-parameter models down to partitioned GPU-right-sized compute instances.
Better together
With roughly 70% of apps and data still residing on premises, the HPE GreenLake platform delivers the cloud experience across your edges, colocations, and data centers. NVIDIA’s pioneering systems, apps, and models combined with the expertise and comprehensive set of computing and infrastructure breakthroughs from HPE can help you unlock the value of AI/ML, VDI, and HPC and lead a data-first modernization. For example, medical imaging pioneer Carestream will transform X-ray systems across the globe using HPE GreenLake for MLOps (read more), and Eni is upgrading its HPC4 system for its Green Data Center (read more).
Don’t miss these HPE sessions at GTC
Attend HPE’s sessions at GTC to discover how to accelerate your AI implementation from edge to cloud with HPE GreenLake cloud services and to network with a global community of developers, researchers, engineers, and innovators.
If you haven’t already, be sure and register at: https://www.nvidia.com/gtc/?ncid=ref-spo-810249
SIMULIVE S42447: From Pilot to Production: Delivering Enterprise-Grade AI as a Service on an Edge-to-Cloud Platform
Glyn Bowden, CTO AI & Data Advisory Services HPE GreenLake, HPE
Wednesday, March 23 8:00AM PT / 4:00PM CET
SIMULIVE S41144: Trustworthy, Explainable AI in Financial Services With Data-Centric Architecture
Aileen Toleikis, Data Analytics Consultant, HPE
Jochen Papenbrock, Financial Services & Technology Developer Relationship Lead EMEA, NVIDIA
Wednesday, March 23 10:00 AM CET / 1:00 AM PT
SIMULIVE S42645: What's the Buzz About Swarm Learning?
Steve Heibein, AI Lead, HPE
Wednesday, March 23 7:00 AM PT / 3:00 PM CET
SIMULIVE S41141: The Underestimated Power: AI & Data Value Creation Within Companies
Thomas Meier, Chief Technologist Data Practice, HPE
Wednesday, March 23 9:00 AM PT / 5:00 PM CET
Session S42455: Scaling Language Training to Trillion-Parameter Models on a GPU Cluster
Evan Sparks, Vice President, AI & HPC, HPE
Session S42462: Optimize Performance and Management of Apache Spark Containerized Workloads
Matthew Morris, Technical Marketing Engineer, HPE
For more information, visit us at HPE GreenLake for AI/ML and Analytics
Flynn Maloy
Hewlett Packard Enterprise
Twitter: @FlynnMaloy
LinkedIn: Flynn Maloy
FlynnMaloy
Flynn Maloy is Vice President of Marketing for HPE GreenLake Cloud Services. Flynn and his team created and launched the HPE GreenLake brand in 2019 and have since grown it into the market leader for on-premises cloud services with over 35% global brand awareness and growing at 82% yr/yr, and with over $4b in contacts. Prior to stepping into his current role, Flynn led the Marketing team for HPE Pointnext Services, the services division of HPE.
