Micron reportedly drafting new $9.6 billion HBM plant in Hiroshima

Micron is reportedly planning a major expansion at its Hiroshima site tied to a $9.6 billion investment and a new high-bandwidth memory plant. Sources say the move is driven by rising demand for Artificial Intelligence-oriented memory and possible subsidies from METI totaling 500 billion yen.

Micron is reportedly preparing a significant expansion of operations at its existing West Japan facility in Hiroshima. The North American computer memory manufacturer currently runs multiple production hubs worldwide and has faced reports of an upcoming exit from China. A Nikkei Asia article cited insiders who link a speculative new plant at the Hiroshima site to a $9.6 billion investment. Present operations at the location focus on DRAM, while the proposed additional factory would specialize in high-bandwidth memory, commonly abbreviated as HBM.

Leaked timelines in the report indicate possible construction starting around May 2026, with initial product shipments — potentially HBM4 and HBM4E — projected to begin in 2028. Expanding Japanese capacity could alleviate pressure on Micron’s Taiwanese factories, which are currently tasked with producing various forms of HBM. Mid-October coverage referenced in the article underscores the broader reshaping of Micron’s global manufacturing footprint that could accompany this project.

The Nikkei Asia piece also points to government incentives as part of the plan. Japan’s Ministry of Economy, Trade and Industry (METI) is said to be preparing significant subsidies, with secretive sources whispering of a 500 billion yen total. The report frames the potential investment and incentives against a backdrop of growing demand for Artificial Intelligence-oriented computer memory, with suppliers still pivoting to meet steep requests from important clients. Details remain speculative and tied to the cited leaks, but the combination of a large capital outlay, a clear timeline, and possible public support would mark a notable expansion of Micron’s presence in Japan.

55

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.