TSMC´s US Supply Chain Could Take Up to a Decade to Establish

TSMC´s expansion in Arizona may take up to ten years to build a supply chain in the US equivalent to its operations in Taiwan and China, due to current production volumes and supplier incentives.

Recent reports from Taiwan media indicate that the Taiwan Semiconductor Manufacturing Company (TSMC) is facing a significant timeline in establishing a robust supply chain in the United States. While TSMC´s suppliers have expressed their willingness to support the company´s US expansion, their current incentives to set up branch offices are limited by the relatively low production volumes in TSMC’s existing American fabs, even after factoring in US tariffs. The consensus among stakeholders suggests it could take up to ten years before TSMC´s US operations are supported by a supply chain comparable to that of its facilities in Taiwan and China.

The Taiwanese government has engaged in discussions with chemical suppliers essential to semiconductor manufacturing, concluding that the present scale of TSMC’s US production does not justify the substantial investment required to create a local supply infrastructure. These stakeholders estimate that establishing a comprehensive US supply chain could take seven to eight years from initiation. However, if TSMC’s production in Arizona scales according to its publicly outlined roadmap, the likelihood increases that more of its suppliers will make the move, setting up their own operations stateside to support the semiconductor giant’s expansion.

An interesting aspect is that TSMC apparently has not prioritized engaging with existing US-based suppliers, including those currently serving Intel. This decision is notable since many of these suppliers provide chemicals and materials similar to those TSMC´s Asia-based partners supply. Possible reasons include uncompetitive pricing from Intel´s suppliers or a preference among TSMC’s established partners to wait until larger volumes are achieved. Over time, as TSMC grows its US fabrication capacity, the semiconductor materials market in the US is expected to become more competitive, benefiting from additional supplier participation and potentially driving innovation and efficiency in the industry.

70

Impact Score

Introducing Mistral 3: open artificial intelligence models

Mistral 3 is a family of open, multimodal and multilingual Artificial Intelligence models that includes three Ministral edge models and a sparse Mistral Large 3 trained with 41B active and 675B total parameters, released under the Apache 2.0 license.

NVIDIA and Mistral Artificial Intelligence partner to accelerate new family of open models

NVIDIA and Mistral Artificial Intelligence announced a partnership to optimize the Mistral 3 family of open-source multilingual, multimodal models across NVIDIA supercomputing and edge platforms. The collaboration highlights Mistral Large 3, a mixture-of-experts model designed to improve efficiency and accuracy for enterprise artificial intelligence deployments starting Tuesday, Dec. 2.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.