Seoul, Apr 19, 2024, IANS

Seoul, April 19 (IANS) South Korean chipmaker SK hynix said on Friday that it has set up a strategic partnership with Taiwan Semiconductor Manufacturing Company (TSMC) aimed at bolstering its capabilities in producing premium high-bandwidth memory (HBM) chips and advanced packaging technologies.

Under the Memorandum of Understanding agreed recently in Taipei, Taiwan, the two semiconductor giants will collaborate in developing sixth-generation HBM4 chips, set for mass production in 2026, according to SK hynix.

Both, SK hynix and TSMC are key customers of Nvidia, a major player in the artificial intelligence (AI) semiconductor market with its graphics processing units, reports Yonhap news agency.

"The collaboration between the global leader in the AI memory space and TSMC, a top global logic foundry, will lead to more innovations in HBM technology," SK hynix said in a statement.

"The collaboration is also expected to enable breakthroughs in memory performance through trilateral collaboration between product design, foundry, and memory provider," it added.

The SK hynix-TSMC collaboration's initial focus will be on enhancing the performance of the base die, which is mounted at the very bottom of the HBM package, SK hynix said.

HBM chips, integral components used for AI computing, have garnered increasing attention with the rise of applications such as generative AI, exemplified by models like ChatGPT.

SK hynix also said it plans to leverage TSMC's advanced logic process for the base die of HBM4, strengthening its functionality to meet diversified customer demands for enhanced performance and power efficiency.

The Korean chipmaker uses a proprietary technology to make base dies up to the current HBM3E.

Moreover, SK hynix and TSMC will also work together to optimise the integration of SK hynix's HBM and TSMC's packaging process, CoWoS technology.

"We expect a strong partnership with TSMC to help accelerate our efforts for open collaboration with our customers and develop the industry's best-performing HBM4," said Justin Kim, head of AI Infra at SK hynix.