-advertisement-

Samsung is reportedly considering upgrading its Taylor fab near Austin from 4nm to 2nm, resulting in delayed equipment orders, according to reports from Tom’s Hardware and ETnews. This decision is expected in Q3 2024.

Originally, Samsung planned to start 4nm mass production at the Taylor plant in the second half of 2024, but this has been postponed to 2025. With the AI hardware competition intensifying and Nvidia, a major client, advancing its GPU platforms, 4nm production is deemed insufficient. Consequently, the Taylor fab's operations may be delayed until 2026, as noted by MySA.

Samsung's press release indicates the company aims to provide comprehensive solutions for AI chip production, with its updated 2nm node featuring backside power delivery expected by 2027. The SF4U, a high-value 4nm variant, is set for mass production in 2025.

In April, the U.S. government announced up to $6.4 billion in subsidies for Samsung to expand its Texas chip production capacity, following previous approvals of up to $8.5 billion for Intel and $6.6 billion for TSMC.

TSMC aims to complete preparations for mass production at its new Arizona fab by the end of this year, starting 4nm production in the first half of 2025. Additionally, TSMC plans to build a third Arizona fab to use 2nm or more advanced technologies, though the construction date is undisclosed.

Intel targets mass production of its 20A and 18A nodes at its Arizona and Ohio plants in 2024 and 2025, respectively.

Editor：Lulu

▼▼▼

Vishay tops out its second factory in Itzehoe

NSIG to spend $1.8 billion on silicon wafer production

Intel starts 3nm chip production in US and Ireland

Onsemi to invest up to $2 billion in Czech semiconductor plant

Taiwan's GlobalWafers receives NT$3.5 billion subsidy for Italian fab

German chipmaker Infineon to set up R&D center in Taiwan