Memory Tiering and the Real Cost of AI | Lenovo & VMware Cloud Foundation

Rising memory costs and unpredictable AI spending are changing how organizations plan their infrastructure. How can IT teams control costs while building for what comes next? Live from VMware Explore 2026, Pete Flecha and Jad El-Zein welcome Scott Patti and David Tareen from Lenovo to discuss memory tiering, hybrid AI, and the infrastructure behind VMware Cloud Foundation. The conversation explores how memory tiering helps address DRAM cost pressures, why tokenomics extends beyond the price of tokens, and how Lenovo Hybrid AI Factory complements VMware AI Factory with infrastructure sizing, stack certification, and deployment services. They also discuss matching AI inference workloads to CPUs and GPUs, bringing enterprise and AI workloads into a familiar operating model, and Lenovo’s Top Choice Express program for getting infrastructure into customers’ hands faster. Chapters: 00:00 Welcome to VMware Explore 2026 00:57 Meet Lenovo and explore the VMware partnership 01:54 Memory tiering and rising infrastructure costs 03:21 How observability builds confidence in memory tiering 04:01 Tokenomics: Understanding the full cost of AI 05:20 Making hybrid AI costs more predictable 06:41 Lenovo Hybrid AI Factory and VMware AI Factory 07:31 Sizing, certification, and deployment services 08:07 Simplifying AI infrastructure and operations 09:20 From metal to model: Matching workloads to infrastructure 10:29 Why Lenovo? Top Choice Express 13:08 Memory tiering configurations and customization 13:49 Pricing commitments and delivery predictability 15:34 Bringing hyperscale experience to enterprise infrastructure 16:24 What customers are talking about at Explore 17:53 Wrap-up Catch all episodes on the VMware Cloud Foundation YouTube channel. #VMwareExplore #VMwareCloudFoundation #Lenovo #PrivateAI #MemoryTiering