Google is moving enterprise AI beyond general-purpose tools and into industry-specific workflows built for regulated sectors such as legal and financial services. Satish Thomas, Vice President, Business and Industry Solutions at Google, joins Six Five to explain how Gemini Enterprise for Industries combines purpose-built skills, ready-to-deploy agents, model choice, and an open partner ecosystem to bring AI into the workflows enterprises already run.

Signal65 President Ryan Shrout and Senior Performance Analyst Mitch Lewis review a new three-level inference infrastructure evaluation of AMD's Instinct MI355X against NVIDIA's HGX B200 with AMD's Andrew Dieckmann and Anush Elangovan. The MI355X's memory bandwidth advantage compounds as context and concurrency scale, delivering up to 2.25x throughput gains and roughly 40 percent lower cost per GPU hour.

Anthropic's "Pace the Frontier" essay set off a week in which nearly every major AI leader, and two heads of state, weighed in on whether the frontier should slow down. Patrick Moorhead and Daniel Newman read the incident as something larger than a safety debate. Whoever sets the pace of the frontier also sets the pace for open-weight models, enterprise custom models, and the trillion-dollar valuations riding on top of them. That contest, not the essay itself, is the through-line of Episode 320 of The Six Five Pod.

Western Digital's Tim Rausch argues that AI's compounding data growth, more than GPU capacity, sets the real ceiling on infrastructure scaling. He joins Matt Kimball at AI InfraSummit 2026 to walk through how total cost of ownership, archived data reuse, and WD's capacity roadmap are reshaping enterprise plans to scale AI over the next several years.

Chad Rikansrud, R&D Software Security Engineer at Broadcom, explains how frontier AI models are finding decades-old mainframe vulnerabilities in seconds, collapsing the complexity advantage that once protected legacy platforms. He outlines why enterprises need faster Security Intelligence (SECINT) ingestion and disciplined human oversight to keep pace with adversaries wielding the same tools.

Physical AI is shifting from lab prototypes to fleets of machines operating reliably in the real world, and Ambarella's Muneyb Minhazuddin joins Six Five On The Road at AI Infra 2026 to explain what that shift demands of silicon, memory and developer workflows. Minhazuddin traces the case back to a 22-year-old chip architecture and forward to a developer cloud built to compress model porting time from months to days, and he identifies enterprise ROI proof points as the biggest hurdle left before physical AI scales into millions of deployed units.

A viral AI extinction claim, a federal advisory naming six Chinese AI labs for industrial-scale distillation, and a $60 billion Qualcomm-AWS silicon pact define a week when enterprise AI risk and opportunity collided head-on. Patrick Moorhead and Daniel Newman weigh in on what's signal and what's noise across AI safety, chip supply deals, and a market rally still betting big on compute.

A Six Five study found an eight-to-one cost gap between running an AI agent through a leading frontier model and running it on modest on-prem hardware. Paul Turner, Chief Product Officer of the VMware Cloud Foundation Division at Broadcom, joins Daniel Newman and Patrick Moorhead at VMware Explore 2026 to explain how VMware's AI Factory turns GPU provisioning, model management, and security into a repeatable path from infrastructure to production AI.

What infrastructure model makes the most sense when AI workloads move from experimentation into production?

In this analyst recap of VMware Explore 2026, Patrick Moorhead and Daniel Newman examine this central question facing enterprise technology leaders and how Broadcom and VMware are positioning private cloud as a critical foundation for bringing enterprise AI into production.