PIM chips integrate the processor with RAM into individual chips. These technological advancements are expected to aid the performance of large-scale AI workloads. In October, Samsung revealed that the company was demonstrating how the performance of its PIM chips could be combined with AMD's graphics processing unit. When the company trained the T5 language model algorithms with this next-generation breakthrough, performance doubled while lowering power consumption in systems without a PIM present.

It now has the world's first large-scale computing system housing 96 AMD Instinct MI100 GPUs combined with 'Processing-In-Memory' PIM technology, which enables enhanced AI algorithms.

Artificial intelligence for science and technology has been accepted in today's world, and is demonstrating widespread support in semiconductor production processes. Samsung is also exploring methods to integrate the knowledge possessed by the company's engineers with inspection data and find use in production.

The Artificial Intelligence Semiconductor Future Technology Conference brings organizations to Korean conference sites, such as the AI Semiconductor Forum, Next Generation Intelligence Semiconductor Foundation, PIM Artificial Intelligence Semiconductor Project Group, as well as demonstrations in the industry, such as DEEPX, Eyenix, ETRI, Furiosa AI, Nextchip, Nepes , Openedges Technology, Sapeon Korea, and Telechips.