Samsung Reveals Advanced Chip Technology to Capture AI Market Share
Samsung Electronics Co. has unveiled a series of technological advancements designed to attract AI chip manufacturers to its production facilities. The announcement was made at the company’s annual foundry forum held at its US chip headquarters in San Jose, California.
Despite being the world’s leading memory-chip maker, Samsung is striving to close the gap with Taiwan Semiconductor Manufacturing Co. (TSMC) in the foundry market, where companies produce chips designed by customers. During the forum, Samsung outlined its chipmaking roadmap and its vision for the AI era.
In the first quarter of this year, Samsung’s share in the foundry market dipped to 11% from 11.3% in the previous quarter, while TSMC’s share rose to 61.7% from 61.2%, according to TrendForce.
The rising demand for components used in AI computing systems has boosted Samsung’s earnings, particularly benefiting its main memory chip division and creating opportunities for outsourcing orders. However, to secure significant commitments from demanding customers like Nvidia Corp., which produces essential AI accelerators, Samsung must demonstrate its production capabilities’ advancement and reliability. The company also faces competition from Intel Corp., which is opening its plants to attract orders from former rivals.
Key to winning orders for AI processors, some of the highest performing and most expensive chips, is the advancement in production technology, typically indicated by smaller transistor dimensions. Samsung introduced an advanced process using backside power delivery network technology, which places power rails on the backside of a silicon wafer. This innovation enhances power, performance, and area efficiency while significantly reducing voltage drop compared to the first-generation 2-nanometer process.
Samsung also highlighted its ability to offer logic, memory, and advanced packaging, positioning itself to rapidly win outsourced semiconductor manufacturing orders for AI-related chips. The company projected that its AI-related customer base would expand fivefold and revenue would increase nine times by 2028. It also announced new production technologies and a layout for future AI-related chips aimed at attracting more customers.
Although Samsung executives did not comment on the status of supplying the latest advanced memory chips to Nvidia or reports of qualification challenges, the company touted its gate-all-around (GAA) technology, crucial for AI products. Samsung plans to mass-produce its second-generation 3-nanometer process later this year and deliver GAA on its upcoming 2-nanometer process. In 2022, Samsung became the first in the industry to start GAA-based 3-nanometer mass production.
The company also confirmed that preparations for its 1.4-nanometer process are progressing smoothly, with performance and yield targets on track for mass production by 2027.