Synopsis: Mike Leach, senior manager for workstation and horizontal solutions at Lenovo, explains how IT infrastructure requirements will need to evolve to build and deploy artificial intelligence (AI) applications.
In this episode of the Techstrong AI video series, host Mike Vizard speaks with Mike Leach, senior manager for workstation and horizontal solutions at Lenovo, about how AI is transforming the role of workstations and endpoints. Leach explains that AI is creating an unprecedented demand for performance, driving the evolution of new types of processors like neural processing units (NPUs) to complement CPUs and GPUs. While high-end GPUs are still essential for training large models, NPUs are emerging as efficient processors for background AI tasks such as inference. This shift is changing how organizations view workstations, making them critical not just for power users, but for anyone who relies on real-time AI-enhanced workflows.
Leach emphasizes that latency is a major challenge for AI applications and a key reason why hybrid approaches—processing AI locally on the device when necessary—are gaining traction. AI models trained in the cloud often need to run at the edge to minimize delays and support real-time decision-making, especially in industries like healthcare, finance, and legal services. Workstations equipped with the right mix of performance components offer the ability to run AI inference locally, ensuring speed and security without relying on a constant cloud connection. This is especially valuable when working with sensitive or proprietary data that can’t be sent off-site.
For IT leaders planning hardware investments, Leach advises embracing AI-capable infrastructure now rather than waiting. He acknowledges that AI-enabled workstations come at a higher upfront cost, but argues that their performance gains lead to better efficiency, lower total cost of ownership, and higher employee productivity. Lenovo helps organizations navigate this complexity by offering curated configurations tailored to specific roles and workloads. Ultimately, Leach stresses that AI isn’t one-size-fits-all; the true business value comes from matching the right AI tools and devices to each user’s unique needs and workflows.
