As artificial intelligence (AI) evolves from training-centric to inference-centric, HPE announced a series of computing and networking products powered by Nvidia graphics processing units (GPUs). It says this will redefine how AI is delivered by moving intelligence to the data and user locations and making the network a trusted fabric that supports real-time experiences.
One of the standout products in this suite is HPE AI Grid, designed to transform distributed AI factories into intelligent AI infrastructure powered by Nvidia. These are said to provide “critical” applications for AI services and use cases that rely on low-latency, real-time connectivity, such as personalization in retail and predictive maintenance in industries such as manufacturing, localized edge inference in healthcare, and carrier-grade AI services.
HPE believes that AI-native applications require a predictable, low-latency, distributed infrastructure. The HPE AI Grid service is part of the Nvidia AI Computing by HPE portfolio and is claimed to deliver predictable, “ultra-low” latency performance at scale for real-time AI services, zero-touch provisioning, and automated security with integrated orchestration.
The end-to-end system is built on the Nvidia reference architecture and is designed to securely connect AI factories and distributed inference clusters across regional and far-end sites. This makes it easy to deploy and operate thousands of distributed inference sites, turning your AI installation into a single intelligent system.
With envisioned use cases in mind, HPE envisioned implementations ranging from retail personalization and predictive maintenance to edge healthcare and carrier-grade AI services, all of which require predictable, ultra-low-latency connectivity. The AI Grid is designed to enable operators to transform existing sites with power and connectivity into a RAN-enabled AI Grid, enabling distributed inference and new services at scale.
HPE AI Grid is said to be differentiated by its ability to work with the AI Grid Reference Architecture to provide service providers with a unified hardware and software stack, as well as full-stack AI servers and AI networks.
HPE AI Grid includes HPE Juniper multicloud routing and coherent optics for predictable long-haul and metro connectivity. Cloud-native and multi-tenant security. Firewall. WAN automation. Orchestration is also available for zero-touch deployment and lifecycle operations. Also included are ProLiant Compute edge and rack servers with Nvidia accelerated computing, including Nvidia RTX PRO 6000 Blackwell GPUs, Nvidia BlueField DPUs, Spectrum-X Ethernet switches, Connect-X SuperNICs, and AI Blueprints for accelerated AI inference.
Neil McRae, Chief Technology and Information Officer at CityFibre, said: “Our customers increasingly expect millisecond responsiveness, low-latency connectivity, and comprehensive security to support their applications and services.
“We are exploring how HPE’s AI Grid, based on Nvidia’s reference architecture, can support distributed AI inference and bring intelligence closer to users and data. By leveraging our fiber network assets, we see the potential to combine high-performance connectivity with intelligent services for our customers.”
said Chris Penrose, Nvidia’s global vice president of communications. “AI Grid brings together geographically distributed AI clusters to optimally place AI workloads and balance performance, cost, and latency across AI factories, regional sites, and the edge.
“We are working with HPE to realize that vision by combining Nvidia’s high-speed computing and networking with HPE’s communications-grade multicloud routing and edge infrastructure to create a single intelligent fabric for distributed inference.”
As part of its AI Grid strategy, carrier Comcast has launched an AI field trial on a distributed network of real-time edge AI inferences to deliver faster, more responsive experiences for the next wave of AI applications. The initial trial tackled several use cases, including leveraging HPE ProLiant servers running Personal AI’s small language models, part of HPE’s Unleash AI partner program, on Nvidia GPUs.
