Samsung has announced it will make 2nm semiconductor chips for Japanese AI firm Preferred Networks and offer advanced packaging for them. This announcement is in line with an earlier report that claimed the same.
Samsung Foundry to make 2nm AI chips for preferred networks
Samsung Foundry has finally received the first client for its 2nm chip fabrication process node. Preferred Networks aims to develop powerful AI accelerators that can offer a lot of computing power for generative AI, and those AI accelerators will use chips fabricated by Samsung Foundry. This chip is designed by GAONCHIPS, a specialized semiconductor chip design firm.
As the South Korean firm recently started offering advanced chip packaging solutions, it will also provide a turnkey semiconductor solution for those 2nm chips for Preferred Networks. A turnkey solution is when a client receives a fully ready-to-use product. Samsung will use its advanced 2.5D packaging technology called Interposer-Cube S (I-Cube S) for 2nm chips.
Samsung was the world's first company to start mass-producing 3nm chips. Its 3nm (or better) process nodes use the Gate All Around (GAA) transistor architecture, which increases drive current and reduces power leakage, leading to more power-efficient chips.
On top of the 2nm GAA process node, Samsung uses the I-Cube S technology, which integrates multiple chips in a single package to improve interconnection speed and reduce the overall size. This should make chips more compact and power-efficient.
Preferred Networks is a Japanese firm that makes advanced hardware and software for AI. It vertically integrates its AI solutions, including AI foundation models, AI accelerators, and supercomputers. Its solutions and products are used in various industries, such as education, entertainment, healthcare, manufacturing, and transportation.
The Japanese firm ranked first three times in the past five years on the Green500 list of supercomputers. Samsung plans to continue its collaboration with Preferred Networks in the future.
Since Samsung announced earlier that its 2nm chips process node will be ready in the second half of 2025, we expect the fabrication of 2nm AI chips for Preferred Networks to go into mass production before the end of next year.
Junichiro Makino, VP and Chief Technology Officer (CTO) of Computing Architecture at Preferred Networks, said, “We are excited to lead in AI accelerator technology with Samsung Electronics’ 2nm GAA process. This solution will significantly support Preferred Networks’ ongoing efforts to build highly energy-efficient, high-performance computing hardware that meets the ever-growing computing demands from generative AI technologies, especially large language models.”
Taejoong Song, Corporate VP and the Head of the Business Development Team at Samsung Foundry, said, “This order is pivotal as it validates Samsung's 2nm GAA process technology and Advanced Package technology as an ideal solution for next-generation AI accelerators. We are committed to closely collaborating with our customers ensuring that the high performance and low power characteristics of our products are fully realized.“
