How HPE GreenLake amplifies the synergies of HPC and AI: An analyst perspective

[Hande_Sahin_Bah]

If it’s true that we’re entering – or already in – the Age of Artificial Intelligence, it’s equally true that high-performance computing (HPC) will be the essential foundation of the new era. In fact, high-performance compute is evolving and becoming even more exciting with the innovations coming from the supercomputing world. Use cases include everything from space exploration to drug discovery, from weather prediction to financial modeling, and from automotive design to systems engineering. Put AI together with supercomputing and the impact can be truly game-changing.

A new report from analyst firm Moor Insights and Strategy, commissioned by HPE and NVIDIA, describes the unique symbiotic link between AI and HPC within a shared infrastructure powered by GPUs: “HPC helps AI process and inference data faster. Conversely, AI helps HPC manage and monitor its processes and workloads more efficiently.” See the report: HPE GreenLake for HPC Delivers an Innovative and Flexible Enterprise HPC/AI Solution. The combination can reduce compute time and lower costs while enabling businesses to “create competitive advantages, improve customer service, spur innovation, and achieve critical business goals.”

That’s great news – and the even better news is that the technology is rapidly becoming far more accessible for businesses through highly scalable, self-service, pay-per-use cloud services. The report explains how HPE GreenLake for high performance computing “brings cloud attributes to HPC, like consumption-based economics and flexible managed infrastructure.”

Read the report to learn:

Why the public cloud is not always the best choice for HPC. While all of the major hyperscalers offer HPC, companies are finding that achieving the expected performance and predictable costs can be a challenge. The report looks at some reasons for this, including egress charges and the cost of premium support.

Why colocation is a powerful option for AI and HPC environments and how HPE delivers this in partnership with industry leaders Equinix, CyrusOne and Digital Realty.

How HPE GreenLake for HPC delivers a fully integrated, modern HPC platform with a cloud-like experience for apps and data.

The report describes HPE GreenLake for HPC as “an excellent solution that enables an enterprise to manage and support an AI/HPC environment now and well into the future.”

I couldn’t agree more. And I’ll just add that, in March, we unveiled a bunch of innovations for this cloud service that will be of interest to any business that’s considering an AI/HPC project. For example, we announced expanded GPU capabilities using NVIDIA A100, A40, and A30 Tensor Core GPUs to help customers advance data-intensive projects across AI, ML and analytics. We also announced HPE Parallel File System Storage for unprecedented throughput, and integration of NVIDIA NVLink for a seamless, high-speed connection between GPUs.

Don’t miss the industry’s biggest supercomputing event, SC22, coming up Nov. 13–18 in Dallas or via an online experience. Visit HPE booths and demos and attend expert sessions to hear about the future of AI with HPE and supercomputing.

Learn more about HPE GreenLake cloud services.

Learn more about HPE GreenLake edge-to-cloud platform and how it opens up new opportunities across your edges, colos, and data centers.