Samsung is now looking forward to setting up to build massive AI factories in South Korea, for which it is considering using over 50,000 Nvidia GPUs.
Samsung Electronics has announced that it is collaborating with Nvidia, aiming to create a giant AI factory in partnership. One of the interesting facts of this collaboration is that Samsung is planning to use over 50,000 Nvidia GPUs to build massive AI factories in South Korea.
Noticeably, these GPUs will be used throughout Samsung’s overall existing manufacturing flow and will also be used in upgrading the development of smartphones, semiconductor chipsets, and robotics-based products.
The key reason behind Samsung’s strategy to use Nvidia’s AI and powerful GPU is to make its semiconductor chips better and faster to produce. AI will be efficient to watch and fix their entire production line in real-time. Although Samsung has already made one part of the chip-making process 20 times faster using Nvidia’s technology. Apart from this, the Korean tech giant has also made Nvidia’s GPU chips in its factories.
Samsung also announced that it is working with Nvidia for the HBM4 chipset, which will be based on its 6th-generation 10nm class DRAM and 4nm logic base die to offer up to 11Gbps data transfer speeds, greatly exceeding JEDEC standards of 8Gbps.
Under the partnership, Samsung will offer GDDR, HBM, and SOCAMM memory chips to build an AI server that uses Nvidia’s GPUs. It will be interesting to see how efficient this partnership will be for both firms.
