Samsung is set to begin mass production of the world’s first sixth-generation HBM4 chips. These advanced chips will play a crucial role in powering next-generation computing systems, specifically in Nvidia’s Vera Rubin AI system, which is expected to launch later this year. The HBM4 chips are designed for high-performance computing in areas such as AI servers and large-scale computing.
A report from South Korea says Samsung could begin shipping these chips to Nvidia as early as next week, around the Lunar New Year. This matches earlier reports about the production schedule. The chips will be used in Nvidia’s Rubin GPUs, which are made for AI servers and large-scale computing. Samsung has been working hard to increase its ability to produce HBM4 chips.
If the report holds true, Samsung will be the first company to mass-produce HBM4 chips, potentially surpassing SK Hynix, the current largest supplier of HBM chips. While Samsung has previously lagged behind SK Hynix in the production of HBM3 and HBM3E chips, it has now caught up with HBM4, positioning itself as a leader in next-generation memory technologies.
Current AI systems like AMD MI350 and Nvidia B200 use HBM3E chips. Micron and SK Hynix are the top suppliers for these, while Samsung has been behind. However, for HBM4, Samsung is expected to become the leading supplier.
Reports suggest that Samsung’s HBM4 chips have already passed Nvidia’s final tests, and Nvidia has placed orders for them. To meet Nvidia’s needs, Samsung has scheduled its production to match Nvidia’s timeline.