The AI Factory Blueprint

Interested in being a guest? Email us at [email protected] is everywhere on stage, but production AI is won in the details. From HPE Discover, we sit down with Jason Schradel, Director of Enterprise Platforms at NVIDIA, to unpack what it really takes to build an “AI factory” that an enterprise can deploy, operate, and scale without turning every upgrade into a science project. If you’re trying to move from AI proofs of concept to real business outcomes, this conversation maps the stack in plain terms. We start with what’s being showcased on the floor and why it matters: next-generation platforms like Vera Rubin and the new Vera CPU, enterprise-ready GPUs including RTX Pro Blackwell Server Edition options, and the networking layer that keeps modern AI workloads moving, from Spectrum-X Ethernet switching to BlueField DPUs. We also talk about how HPE systems, storage, and the private cloud experience come together with NVIDIA accelerated computing and NVIDIA AI software to form a repeatable blueprint for enterprise AI infrastructure. From there, we zoom out to the go-to-market reality: global customers, real deployments across industries like healthcare, manufacturing, financial services, and telecom, plus the growing role of ISVs and partner ecosystems in making AI usable for specific workflows. Jason also shares what he’s watching on the roadmap, especially agentic AI and the importance of confidential computing to protect sensitive data and model weights as hybrid cloud AI becomes the norm. If you’re planning an enterprise AI strategy, you’ll leave with a clearer view of the components that matter most and the tradeoffs you can’t ignore. Subscribe for more conversations like this, share this episode with a teammate building your AI platform, and leave a review. What part of the AI factory stack feels hardest to get right right now?Support the showMore at https://linktr.ee/EvanKirstel