NVIDIA has reached a cooperation with SK Hynix to jointly develop a new generation of memory chips for artificial intelligence, which is a great benefit for this leading Korean high-end semiconductor company. The two companies said in a statement that they have signed a multi-year agreement to jointly promote chip design and manufacturing business. NVIDIA will help partners expand new business areas, covering infrastructure, physical artificial intelligence, and storage products supporting NVIDIA's flagship acceleration chip Vera Rubin.
Nvidia CEO Jensen Huang confirmed for the first time last week that the company has approved Samsung Electronics, SK Hynix and Micron Technology to supply HBM4 memory. This product is an indispensable core component of the Vera Rubin chip. These three companies dominate the global storage market and are fiercely competing for this high-margin business.
Huang Jen-Hsun revealed at the Taipei Computex that the Vera Rubin chip has now entered full mass production and is expected to be officially delivered in the third quarter of this year. The new computing system is based on NVIDIA Vera series central processors and Rubin graphics core clusters, and each server is equipped with HBM4 memory with several terabytes of capacity.