HPE Companies, in the meantime, is offering a brand new providing that leverages Nvidia NeMo frameworks to create what HPE calls “digital avatar assistants” designed to ship AI-based buyer transactions for purposes for good cities, retail, manufacturing, well being care, banking, finance and training.
There are a number of new {hardware} merchandise obtainable from HPE:
- HPE ProLiant Compute XD685 now helps eight Nvidia Blackwell Extremely (B300 GPUs) in a 5U, direct-liquid cooled HGX chassis. The XD685 is optimized for AI service suppliers and neoclouds, in addition to formidable enterprises and mannequin builders that want to face up massive, safe, and validated AI clusters.
- Nvidia GB300 NVL72 by HPE, optimized for AI fashions of greater than 1 trillion parameters, is now obtainable to order and anticipated to ship December 2025.
- HPE ProLiant Compute DL380 Gen12 Server Premier Answer for Azure Native, designed to supply Azure providers on a buyer’s community in addition to AI providers.
- Future fashions of HPE servers will supply help for Nvidia ConnectX-9 SuperNIC and the Nvidia BlueField-4 DPU, introduced yesterday.
HPE cited a 2025 Architecting an AI Benefit report that discovered almost 60% of organizations have fragmented AI objectives and methods, and the same proportion missing complete knowledge administration for AI.
The 2 corporations are collaborating to handle these challenges by offering a collection of options for AI factories, supporting a spread of customers from builders to sovereign AI cloud builders. The aim is to allow organizations to develop a holistic AI infrastructure technique.
Flex and Nvidia collaborate on giga-scale AI Factories
Nvidia struck a take care of manufacturing developer Flex to develop modular knowledge heart methods designed to fulfill the surging demand for AI infrastructure throughout the US. The collaboration is constructed on Flex’s superior manufacturing capabilities and international footprint to ship high-performance, energy-efficient AI factories at scale.
Flex is utilizing Nvidia’s cuOpt for capability planning, route simulations, and course of optimization. cuOpt is a GPU-accelerated optimization engine developed by Nvidia for fixing massive decision-optimization issues similar to route planning and scheduling for car fleets.
