South Korean chipmakers Samsung Electronics and SK hynix will showcase their latest chips for artificial intelligence applications and supercomputing at Supercomputing 24, the largest international conference on high-performance computing, next week.

The annual event which first kicked off in 1988 and is organized by Association for Computing Machinery and the IEEE Computer Society, will take place at the Georgia World Congress Center in Atlanta, from Sunday to Nov. 22.

The Korean chipmakers will join 345 other global tech firms such Nvidia, Microsoft and Google to share their industry outlook and showcase their chips at the event.

Samsung, the world’s largest memory chipmaker, is expected to exhibit its latest products utilizing the Compute Express Link, according to industry sources.

CXL is a next-generation interface that adds efficiency to accelerators, DRAM and storage devices used with central processing units in high-performance server systems. It can be comparable to merging wide roads, enabling the efficient transfer of large volumes of data, according to a Samsung official.

At the event, Samsung is expected to showcase the CXL Memory Module-DRAM (CMM) and CMM-H, which combines the CMM with NAND flash memory chip.

The company is also expected to promote its High Bandwidth Memory chips, the most advanced DRAM chip that enhances AI processing in graphic processing units. The chip giant is expected to start supply the fifth-generation 12-layer HBM3E to customers in the October-December period.

SK hynix, the world's second-largest memory chip maker and the dominant supplier of the HBMs, is expected to exhibit the AI chips along with AI training solutions utilizing CXL.

The chipmaker is the main supplier of the high-value HBM3E chips to Nvidia, the largest GPU maker in the US, and is currently leading the advanced DRAM market taking over 52 percent share.

SK hynix is also expected to introduce Accelerator-in-Memory based Accelerator (AiMX), an accelerator card that specializes in large language models, based on its GDDR6-AiM chip. The high-performance memory product provides high bandwidth and outstanding energy efficiency, according to the company.

Supercomputing is a form of high-performance computing that determines or calculates by using a powerful computer, reducing overall time to solution.

By: DocMemory
Copyright © 2023 CST, Inc. All Rights Reserved