According to Preferred Networks, by leveraging Samsung’s foundry and advanced packaging products, it will look to develop powerful AI accelerators that will address the demand for computing power driven by generative AI.
Since starting mass production of the industry’s first 3nm process node applying Gate-All-Around (GAA) transistor architecture, Samsung has won a number of orders for the 2nm process with further upgrades in performance and power efficiency.
The cooperation with Preferred Networks marks the first achievement for Japanese companies in the field of large-sized heterogeneous integrated package technologies.
The 2.5D Advanced Package I-Cube S technology included in the turnkey solutions, is a heterogeneous integration package technology, with multiple chips in one package to enhance inter-connection speed and reduce package size. The use of the silicon interposer (Si-interposer) is crucial in achieving ultra-fine redistribution layer (RDL) and stabilizing power integrity for optimal semiconductor performance. GAONCHIPS, a specialised system semiconductor development company, designed the chip.
"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,” said Junichiro Makino, VP and Chief Technology Officer (CTO) of Computing Architecture at Preferred Networks.
"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," explained Taejoong Song, Corporate VP and the head of Foundry Business Development Team at Samsung Electronics. "We are committed to closely collaborating with our customers ensuring that the high performance and low power characteristics of our products are fully realised."
Preferred Networks is a developer of advanced software and hardware technologies by vertically integrating the AI value chain from chips to supercomputers and generative AI foundation models.
