As tech giants commit billions to GPU clusters and cloud compute sells out years in advance, are we witnessing sustainable exponential innovation or an impending infrastructure bubble? In this episode recorded ahead of AI Infrastructure Field Day 6, host Alastair Cooke is joined by panelists Frederic Van Haren, Brandon Whichard, and James Urquhart to break down how the lines are blurring between AI training and inference, hyperscalers and neo-clouds, and deterministic systems versus generative models. Together, they analyze the financial realities behind massive capex commitments, investigate whether enterprise revenue can justify insatiable hardware demand, and discuss what happens when AI models finally become “good enough” for everyday business needs. This and more on the Tech Field Day Podcast, part of the Futurum Media Podcast Network.
Panelists
Alastair is a Tech Field Day event lead at the Futurum group, specializing in Cloud, DevOps, and Edge.
Brandon Whichard is the co-host of Software Defined Talk and founder of Failover Media, a B2B podcast network reaching cloud architects, platform engineers, and IT decision-makers across three shows and 20M+ combined downloads. He brings 20+ years of product management and marketing experience from Sun Microsystems, IBM, CA Technologies, DXC Technology, and several startups.
Frederic is the CTO at HighFens Inc (active in the HPC & AI market). He spent more than a decade building large HPC & AI environments from the ground up and is frequently invited to speak at events to provide his vision on AI, Big Data and HPC.
Father, husband, technologist, communicator. VMware Tanzu. Former blogger (CNET, GigaOm), author Flow Architectures (O’Reilly)