AMD and HPE Expand Collaboration to Advance Open Rack-Scale AI Infrastructure

Core Insights - AMD has announced an expanded collaboration with HPE to accelerate the development of open, scalable AI infrastructure using AMD's advanced computing technologies [2][4] - The new AMD "Helios" rack-scale AI architecture will be adopted by HPE, integrating specialized networking solutions for enhanced performance [7][9] - The "Helios" platform is designed to deliver up to 2.9 exaFLOPS of FP4 performance per rack, utilizing AMD's latest hardware and software technologies [5][6] Collaboration Details - HPE will leverage AMD's "Helios" architecture to provide cloud service providers with faster deployment and greater flexibility in scaling AI computing [4][9] - The partnership aims to redefine high-performance computing capabilities, focusing on open standards to foster innovation [4][8] Technical Specifications - The "Helios" platform combines AMD EPYC CPUs, AMD Instinct GPUs, and advanced networking technologies to create a cohesive system optimized for AI workloads [3][5] - The system is built on the OCP Open Rack Wide design, facilitating streamlined deployment for demanding AI applications [6][7] Future Developments - HPE plans to offer the AMD "Helios" AI Rack-Scale Architecture globally in 2026, enhancing AI capabilities for various industries [7][9] - A new supercomputer named "Herder," powered by AMD technologies, will be deployed at the High-Performance Computing Center Stuttgart in Germany, expected to enhance HPC and AI research [8][10]