John Hampton, AMD | Red Hat Summit 2026
summary
Enterprises face two concrete problems: aging fleets that can shrink from 1,000 servers to 150 to free budget and power, and exploding AI compute costs measured as tokens per dollar and tokens per watt. AMD maps workloads to EPYC CPUs, the new PCIe GPU MI350P (available with Red Hat 3.5 EA), and scale-up Kyrios systems to optimize cost and power. Only single-digit percentages of firms reach production today; AMD reports 150+ internal AI use cases that boost engineer productivity rather than replace them.
AI-written summary · about
chapters (9)
- 0:00Welcome and introduction
- 1:29Enterprise AI infrastructure challenges
- 3:36AMD Red Hat partnership benefits
- 5:02Understanding tokenomics in AI
- 7:10New AMD Instinct MI350P launch
- 8:58AI leaders vs laggards differences
- 11:03AI replacing jobs misconceptions
- 12:48AI impact on leadership approach
- 14:32Future vision for Red Hat Summit 2027
