en.Wedoany.com Reported - At the HPE Discover conference, Jason Schradel, Director of Enterprise Platforms at NVIDIA, discussed with HPE the technical elements required to build enterprise AI factories, focusing on transitioning enterprises from AI proof-of-concept stages to deployable, operable, and scalable production environments, while outlining key aspects of the overall technology stack.

The conversation began with booth highlights, covering next-generation platforms such as Vera Rubin, Vera CPU, enterprise-grade GPUs (including the RTX Pro Blackwell server edition option), and the networking layer supporting continuous operation of modern AI workloads—from Spectrum-X Ethernet switching to BlueField DPUs. The integration of HPE systems, storage, and private cloud experiences with NVIDIA accelerated computing and AI software was seen as a path to forming a replicable blueprint for enterprise AI infrastructure.

The discussion also addressed market realities, including real-world deployments by global customers in industries such as healthcare, manufacturing, financial services, and telecommunications, as well as the growing role of independent software vendors (ISVs) and partner ecosystems in adapting AI workflows. Jason Schradel shared his focus on the technology roadmap, particularly agentic AI and the importance of confidential computing—which protects sensitive data and model weights when hybrid cloud AI becomes the norm.