Samsung has entered mass production of Nvidia’s inference chip Groq 3 LPU, marking a significant step in its foundry business recovery.

Samsung has entered mass production of Nvidia’s inference chip Groq 3 LPU, marking a significant step in its foundry business recovery.

Securing volume fabrication for advanced inference silicon validates Samsung Foundry’s process node competitiveness against TSMC and expands the multi-source supply chain for AI accelerators.