industry desk
立場別 · 1 takes across the edition
TechTimes · United States · Samsung bets on a memory-plus-foundry dual engine as Tesla and Nvidia knock
Reports the June 8 Seoul bilateral: HBM4E samples already at Nvidia, a roadmap through HBM5, and talks for Samsung Foundry to build Nvidia's next-gen Groq LP40 chip. Frames it as Samsung trying to convert foundry access into the HBM share it lost to SK Hynix in the Rubin generation.
“Samsung pursues Nvidia HBM4E supply and LP40 foundry work after a Seoul bilateral, trying to convert foundry access into lost HBM share.”