Industrial AI Podcast

Industrial AI Hardware: "CUDA was the last 15 years"

Sep 17, 2025
Jens Stapelfeldt, Global AI Lead for AMD's University Program, discusses AMD's innovative approach to industrial AI. With decades of experience in the semiconductor industry, he outlines the shift from proprietary ecosystems to AMD’s open platforms. Jens highlights the importance of ROCm and strategic partnerships in enhancing edge-to-cloud solutions. He emphasizes how diverse hardware options and open-source principles are reshaping industrial use cases, making AMD a versatile player in the AI landscape.
Ask episode
AI Snips
Chapters
Transcript
Episode notes
ANECDOTE

Career Path Into AMD University Program

  • Jens has 30 years in semiconductors and joined AMD via the Xilinx acquisition over a decade ago.
  • He studied AI startups in a part-time MBA and now works with universities and startups in AMD's University Program.
ADVICE

Use AMD University Program Resources

  • Universities and startups should leverage AMD's donations, training, and funded research via the AMD University Program.
  • Jens encourages using AMD's prepared research material and resources to accelerate academic AI work.
INSIGHT

Open, Competitive, Multi-Vendor Strategy

  • AMD positions its advantage as threefold: competitive GPU performance, an open software stack, and broad hardware choice.
  • Jens argues this combination enables multi-vendor, cost-optimized AI beyond the CUDA era.
Get the Snipd Podcast app to discover more snips from this episode
Get the app