HPE and NVIDIA Collaborate on GenAI

Credit: HPE and NVIDIA

The HPE Discover 2024 conference is currently in full swing, and the keynote address from Hewlett-Packard Enterprise (HPE) CEO Antonio Neri on Tuesday, June 18, was an unforgettable event. Other than being the first business keynote hosted at the Sphere near Las Vegas, Nevada, the keynote also unveiled an exciting new collaboration between NVIDIA and HPE.

Specifically, this partnership has culminated in the NVIDIA AI Computing by HPE portfolio of co-developed AI solutions to help enterprises accelerate the adoption of generative AI. By tightly integrating their technologies, NVIDIA’s leading AI technologies will be combined with the HPE partner network to bring the power of AI to enterprise customers on a massive scale.

A Keynote to Remember

The keynote itself from Neri was exciting in its own right, as it took full advantage of the enormous dome-like screen at the Sphere event center. Beginning with a host of gorgeous video clips of the natural world, Neri eventually walked onstage to thunderous applause. His first words set the tone for the rest of the speech:

“Big moments require big venues,” Neri said. “Welcome to my living room.”

The Sphere seems an appropriate venue for such a massive announcement. Credit: HPE

After an introductory talk discussing HPE’s history, Neri spoke on the “potential and promise of AI and catapult the enterprise of today and tomorrow to new heights.” This eventually led to Neri stating that NVIDIA has long been a “visionary partner who shares our purpose and commitment to innovation.” Following a list of past HPE-NVIDIA collaborations, Neri announced NVIDIA AI Computing by HPE and brought Jensen Huang onto the stage.

In his trademark black leather jacket, Huang oozed enthusiasm as he jogged onstage and almost immediately shouted “Go HP!” When looking at the collaboration, his excitement is understandable.

The NVIDIA AI Computing by HPE portfolio is packed full of useful tools, but one of the more key offerings is the HPE Private Cloud AI. Hailed by HPE as a “turnkey solution for every industry,” this cloud-based tool combines NVIDIA AI computing, networking, and software with HPE’s AI storage, compute, and the HPE GreenLake cloud platform.

The company offers support for inference, fine-tuning, and RAG AI workloads that utilize proprietary data. Potential buyers can expect enterprise control for data privacy, security, transparency, and governance requirements. The cloud experience has ITOps and AIOps capabilities to increase productivity, and the tool offers a fast path to consume energy flexibly to allow companies to meet future AI opportunities.

HPE Private Cloud AI provides a fully integrated AI infrastructure stack that includes NVIDIA Spectrum-X Ethernet networking, HPE GreenLake for file storage, and HPE ProLiant servers with support for NVIDIA L40S, NVIDIA H100 NVL Tensor Core GPUs, and the NVIDIA GH200 NVL2 platform to deliver optimal performance for the AI and data software stack.

Additionally, HPE is adding support for NVIDIA’s latest GPUs, CPUs, and Superchips. The HPE Cray XD670 supports eight NVIDIA H200 NVL Tensor Core GPUs, ideal for large language model (LLM) builders. The HPE ProLiant DL384 Gen12 server with NVIDIA GH200 NVL2 is tailored for LLM consumers using larger models or retrieval-augmented generation (RAG). The HPE ProLiant DL380a Gen12 server supports up to eight H200 GPUs, providing flexibility for scaling generative AI workloads. Additionally, HPE plans to support NVIDIA’s upcoming GB200 NVL72/NVL2, Blackwell, Rubin, and Vera architectures.

Observability and AIOps are also provided to all HPE products and services through the integration of OpsRamp’s IT operations with HPE GreenLake cloud. The whole NVIDIA accelerated computing stack, comprising NVIDIA NIM and AI software, NVIDIA Tensor Core GPUs and AI clusters, NVIDIA Quantum InfiniBand and NVIDIA Spectrum Ethernet switches, is now observable with OpsRamp. IT managers may monitor their workloads and AI infrastructure in hybrid and multi-cloud settings by gaining insights to spot irregularities.

Partners will begin quoting customers for HPE Private Cloud AI on July 8 and shipping begins in September.

Comparisons to the Dell Collaboration

Those who keep abreast of NVIDIA news may see some similarities between this partnership and a recently announced collaboration between NVIDIA and Dell. Both involve partnerships between enterprise infrastructure companies and NVIDIA to deliver integrated and optimized generative AI solutions for enterprises.

Credit: Dell Technologies and NVIDIA

Dell and NVIDIA’s partnership is meant to expand upon the Dell Generative AI Solutions portfolio, which includes the Dell AI Factory with NVIDIA. This is an integrated, end-to-end AI enterprise solution that combines Dell’s compute, storage, software, and services with NVIDIA’s AI infrastructure and software suite to support the full generative AI lifecycle. It is also available via traditional channels as well as Dell Apex.

This is one key difference between Dell AI Factory and HPE Private Cloud AI, as HPE is partnering with system integrators while Dell is leveraging its traditional channels.

Dell is also working to support new NVIDIA GPU models. The NVIDIA B200 Tensor Core GPU, which is anticipated to provide up to 15 times better AI inference performance and a reduced total cost of ownership, is one of the new NVIDIA GPU models that Dell PowerEdge XE9680 servers will support. Other GPUs based on the NVIDIA Blackwell architecture, H200 Tensor Core GPUs, and the NVIDIA Quantum-2 InfiniBand and Spectrum-X Ethernet networking platforms will also be supported by the Dell PowerEdge servers.

NVIDIA has established itself as the “kingmaker” in the AI industry, and its decision to deeply integrate its stack with these partners is a powerful endorsement. Collaborations such as these will be vital for enterprises to truly unlock the AI’s transformative potential.

Related