Episode
The Quiet War Over AI Chip Memory Bandwidth
- Published
- Jun 23, 2026
- Duration seconds
- 439
- Processing state
not_requested
Actions
POST https://stenobird.com/v1/public/podcasts/the-ai-podcast-with-fexingo-artificial-intelligence-machine-learning-and-modern-ai-models-7872011/episodes/the-quiet-war-over-ai-chip-memory-bandwidth/transcription-requests
Idempotently request low-priority transcript generation for this episode.GET https://stenobird.com/podcast/the-ai-podcast-with-fexingo-artificial-intelligence-machine-learning-and-modern-ai-models-7872011/the-quiet-war-over-ai-chip-memory-bandwidth.md
Read the agent-friendly Markdown representation of this episode resource.
Summary
Everyone focuses on compute—teraflops, H100s, GB200s. But the real bottleneck in AI inference right now is memory bandwidth: how fast the chip can feed data to those compute cores. Lucas and Luna dive into HBM3E, the race between SK hynix, Samsung, and Micron, and why ASML and chip equipment stocks may be the real winners. They look at the numbers: AMD up just 0.8% this week while Micron surged 11.3% and ASML rose 1.9%. The episode explores why bandwidth matters more than raw FLOPS for inference workloads, how HBM is evolving, and what this means for long-term AI infrastructure investing. #AI #ArtificialIntelligence #Technology #Investing #Semiconductors #Memory #HBM #Micron #ASML #AMD #NVDA #NVIDIA #Inference #ChipDesign #TechStocks #FexingoBusiness #BusinessPodcast #AIHardware Keep every episode free: buymeacoffee.com/fexingo