AI Unlocked
1823864 Members
4367 Online
109666 Solutions
New Article ๎ฅ‚
HPE_Experts

HPE ProLiant Compute DL380a Gen12 expands performance with NVIDIA H200 NVL

Explore the HPE ProLiant DL380a Gen12, optimized for enterprise AI with NVIDIA H200 NVL, offering 1.7x faster inference, enhanced memory, and energy-efficient performance for cutting-edge innovation.

HPE-ProLiant-NVIDIA-NVL.pngHewlett Packard Enterprise (HPE) introduced the HPE ProLiant Compute DL380a Gen12 server at HPE Discover in June 2024 as part of our NVIDIA AI Computing by HPE announcement, aimed at optimizing AI fine-tuning and inferencing for enterprise organizations. This highly scalable, GPU-accelerated server was designed for enterprise-level reliability, equipped with up to eight power supplies, with six of those dedicated to powering GPUs, enhancing overall performance and stability during intensive computing tasks.

As enterprise organizations navigate AI deployments and infrastructure, it is important that they have architecture flexibility and choice. PCIe GPU options can help them meet the need for data centers that require lower-power, air-cooled rack designs while ensuring they can tackle complex AI workloads. During the Supercomputing 2024 conference, NVIDIA announced the availability of NVIDIA H200 NVL, part of the NVIDIA Hopper architecture, which delivers higher performance and greater energy efficiency in PCIe form factor.

While the HPE ProLiant Compute DL380a Gen12 has been available to order for several weeks now, today we are excited to highlight HPEโ€™s support for the new NVIDIA H200 NVL. The HPE ProLiant Compute DL380a Gen12 with the new NVIDIA H200 NVL is featured on the expo floor at both Supercomputing 2024 and the HPE Discover Barcelona conference this week.

We are excited about the performance and value that the NVIDIA H200 NVL  provides, including:

  • 1.7x faster large language model inference over NVIDIA H100 NVL
  • 1.5x memory increase and 1.2x bandwidth increase over NVIDIA H100 NVL
  • 7x faster GPU-to-GPU communication over PCIe Gen 5. With up to 4 PCIe GPUs connected via NVIDIA NVLink, NVIDIA H200 NVL delivers higher performance for large language model inference and fine-tuning
  • A five-year subscription for NVIDIA AI Enterprise, a cloud-native software platform for the development and deployment of production AI

Empowering enterprise organizations to advance their AI initiatives, while packing more computing powers into smaller spaces, leveraging their existing racks, with faster and more efficient operations, is exciting. HPE is committed to helping you get the most out of your AI innovation, with the flexibility you need. The HPE ProLiant Compute DL380a with the new NVIDIA H200 NVL is now orderable. 

To learn more, visit https://www.hpe.com/us/en/compute/proliant-dl380a-gen12.html


 

Aaron Lamond.pngMeet HPE Blogger Aaron Lamond, Worldwide Product Marketing, AI Solutions, HPE Compute

xtensive experience in engineering and sales. Aaron is based in New Zealand.  Follow him on LinkedIn.
 
 
About the Author

HPE_Experts

Our team of Hewlett Packard Enterprise experts helps you learn more about technology topics related to key industries and workloads.