Previously, we learned that Intel had partnered with Terafab, with Tesla becoming Intel’s first major customer for node 14A of Elon Musk’s Terafab AI chip complex in Austin. However, the details of this agreement remain unclear, including whether Terafab will obtain a license for the PDK of node 14A. Although TSMC and Intel may coexist within the Terafab project, it is important to note that TSMC may not import its latest semiconductor technology outside of Taiwan, keeping its most advanced nodes at its headquarters, while Terafab remains a secondary project. Perhaps Intel and TSMC will offset each other where the technology needs to be balanced, but that remains undetermined.
Terafab’s goal is to bring the entire chip manufacturing process under one roof. The factory aims to integrate multiple stages of semiconductor production at a single site, including logic manufacturing, memory, packaging, testing and mask production. This setup is unusual because these steps are typically spread across multiple facilities and specialized companies.
Additionally, simply investing more funds does not guarantee a solution. Building a competitive factory from scratch requires years of execution, including sourcing parts from companies like ASML, Lam Research, KLA Corporation and Tokyo Electron. These companies often receive orders years in advance before delivering machines to TSMC, Samsung or Intel. Thus, it will take years to create a supply chain of equipment, raw materials and other necessary components. Terafab has reportedly already started ordering tools for its manufacturing process from these companies, so we can expect the first silicon in a few years.
Gn bussni
