Nvidia (NVDA) CEO Huang launches new Groq LPU product and Vera Rubin opens agentic AI Frontier; Launches bluefield-4 stx storage architecture with broad industry adoption

• Unveils new Groq 3 LPX AI inference chip based on startup's tech.
• New Groq-based server will have eight Groq chips, which are third gen of chip.
• Co. is in full production of Co-packaged Optical chips for Spectrum X chips with TSMC.
• CPU for sure a multi-billion dollar business.
• Thanks Samsung, who manufactures Grok 3 chip, and in production with Grok chip and they'll ship it in H2, probably Q3 timeframe.
• Copper is still going to be important, and they are going to be scaling up Optical.
• Next gen Feynman design will have HBM memory.

Nvidia's launch of the Groq 3 LPU product signifies a major move in the AI and semiconductor space, enhancing its competitive position in the booming AI market. This could accelerate the adoption of their technologies across industries, potentially translating into substantial revenue growth and strong demand for advanced computing solutions. The collaboration with key partners like TSMC and Samsung underlines their commitment to scaling production effectively.