HPE accelerates AI deployments with first AMD “Helios” AI rack-scale architecture with open, scale-up networking built with Broadcom

Purpose-built HPE Juniper Networking scale-up switch and software are first in industry to supercharge AI training and inference performance using standards-based Ethernet

In this article

• HPE announces AMD “Helios” AI rack-scale solution with industry-first scale-up Ethernet networking in collaboration with Broadcom.
• HPE will be one of the first companies to offer the single turnkey rack capable of trillion parameter AI training and high-volume inferencing, delivering 260 terabytes of aggregate scale-up bandwidth and 2.9 AI exaflops of FP4 performance.
• The AMD “Helios” AI rack-scale architecture is an open, integrated platform for large-scale AI workloads, putting industry standards at the forefront of the design.

BARCELONA--(BUSINESS WIRE)--HPE Discover Barcelona 2025 – HPE (NYSE:HPE) today announced it is speeding up deployment of at-scale artificial intelligence (AI) training and inferencing for cloud service providers (CSPs), including neoclouds, by offering the first AMD “Helios” AI rack-scale architecture with integrated scale-up Ethernet networking. The solution, leveraging purpose-built HPE Juniper Networking hardware and software along with Broadcom’s Tomahawk 6 networking chip, is based on the open Ultra Accelerator Link over Ethernet (UALoE) standard and capable of supporting the traffic from trillion parameter model training, high inference throughput, and massive model sizes. Delivered by an HPE Services team with extensive experience in direct liquid-cooling infrastructure and exascale installations, the AMD “Helios” AI rack-scale solution will offer customers flexibility, interoperability, energy efficiency, and faster deployments amidst greater industry demand for AI compute capacity.

“For more than a decade, HPE and AMD have pushed the boundaries of supercomputing, delivering multiple exascale-class systems and championing open standards that accelerate innovation,” said Antonio Neri, president and CEO at HPE. “With the introduction of the new AMD ‘Helios’ and our purpose-built HPE scale-up networking solution, we are providing our cloud service provider customers with faster deployments, greater flexibility, and reduced risk in how they scale AI computing in their businesses.”

“HPE has been an exceptional long-term partner to AMD, working with us to redefine what is possible in high-performance computing,” said Dr. Lisa Su, chair and CEO at AMD. “With ‘Helios’, we’re taking that collaboration further, bringing together the full stack of AMD compute technologies and HPE’s system innovation to deliver an open, rack-scale AI platform that drives new levels of efficiency, scalability, and breakthrough performance for our customers in the AI era.”

Accelerating AI deployments with a single turnkey rack with industry-first Ethernet-based solution

Delivering a leap forward in rack-scale AI performance, HPE will be one of the first companies to offer the AMD "Helios" rack-scale solution. It is based on specifications from the Open Compute Project (OCP) that are optimized for power delivery, advanced liquid cooling, and high serviceability as needed for next-generation AI systems. Additional features of the integrated solution include:

• Rack-scale performance: Connecting 72 AMD Instinct™ MI455X GPUs per rack, the solution delivers 260 terabytes/second of aggregated scale-up bandwidth and up to 2.9 AI exaflops of FP4 performance to accelerate workloads and support massive AI models. The system provides 31 terabytes of fourth generation high bandwidth memory (HBM4) and 1.4 petabytes/second of memory bandwidth, delivering uncompromised capacity and data movement for the most demanding AI and HPC workloads.
• Industry-first standards-based Ethernet scale-up switch and software: HPE is extending its “networks for AI” solutions with a scale-up Ethernet switch and software designed specifically for the AMD “Helios” AI rack-scale architecture. The new solution is the first scale-up switch to deliver optimized performance for AI workloads over standard Ethernet, developed in collaboration with Broadcom. The switch leverages HPE’s AI-native automation and assurance capabilities to simplify network operations for faster deployment and overall cost savings. The new HPE scale-up solution complements HPE’s existing scale-out and scale-across offerings, delivering a complete “networks for AI” portfolio.
• Open standards and innovation: Built on OCP’s Open Rack Wide (ORW) specifications, the double-wide design of AMD “Helios” AI rack-scale architecture offers power and energy efficient liquid cooling in a practical design optimized for ongoing maintenance. The AMD “Helios” AI rack-scale architecture also leverages the open source AMD ROCm™ software and AMD Pensando™ networking technology to speed innovation and help lower total cost of ownership. Based on an open and proven communications standard, the Ethernet-based HPE Juniper Networking switch minimizes vendor lock-in and enables swifter feature updates.

Broadcom is working with HPE to develop the HPE Juniper Networking scale-up switch for the AMD “Helios” AI rack-scale architecture, leveraging the two companies’ longstanding collaboration and commitment to open Ethernet technology for AI data center networks. This effort aims to deliver scalable, efficient, and high-performance networking solutions tailored for modern AI workloads.

“Broadcom is proud to take part in this collaboration to advance open, Ethernet-based AI infrastructure for scale-up,” said Hock E. Tan, president and CEO at Broadcom. “Our high-performance silicon delivers industry-leading ultra-low latency, massive performance, and lossless networking with the scalability and efficiency modern AI workloads require. Together with HPE and AMD, we are enabling customers to build powerful AI data centers with standard Ethernet, maximizing choice and flexibility while delivering exceptional scale.”

Availability

HPE will offer the AMD “Helios” AI rack-scale solution worldwide in 2026.

About HPE

HPE (NYSE: HPE) is a leader in essential enterprise technology, bringing together the power of AI, cloud, and networking to help organizations achieve more. As pioneers of possibility, our innovation and expertise advance the way people live and work. We empower our customers across industries to optimize operational performance, transform data into foresight, and maximize their impact. Unlock your boldest ambitions with HPE. Discover more at www.hpe.com.