Japan is pushing harder into advanced chip manufacturing, and Fujitsu now plans to work with Rapidus to build what could become one of the world’s first 1.4nm AI processors, with a clear focus on strengthening domestic semiconductor capabilities while reducing reliance on overseas production.
The project centers on a dedicated neural processing unit designed for artificial intelligence workloads, and Fujitsu intends to pair this chip with its high-end Monaka processors, creating a tightly integrated platform aimed at next-generation supercomputing systems such as Fugaku NEXT.
Fujitsu targets advanced AI compute with Monaka platform
Fujitsu’s strategy builds around its Monaka architecture, which already targets high-performance computing with features such as up to 144 cores per socket, a 3D chiplet design, and support for PCIe 6.0 and CXL 3.0, and when combined with a 1.4nm NPU, the company aims to deliver significantly higher compute density and efficiency for AI-heavy workloads.
This pairing reflects a broader shift where companies integrate CPUs and AI accelerators more closely, and Fujitsu’s approach signals a move toward tightly coupled systems designed specifically for large-scale data processing and machine learning tasks.
Rapidus to handle manufacturing with government backing
Nikkei Asia reports that Fujitsu will rely on Rapidus for manufacturing, while Japan’s Ministry of Economy, Trade and Industry is expected to support the project financially as part of a wider push to build a domestic chip ecosystem.
“Manufacturing will be commissioned to Rapidus, a Japanese semiconductor company seeking to mass-produce cutting-edge chips. Japan’s Ministry of Economy, Trade and Industry is expected to cover part of the development cost. The effort will be a first step in responding to global moves to build AI capabilities amid economic security concerns.” — Nikkei Asia
This focus on “economic security concerns” shows how governments now treat chip production as a strategic priority, especially as AI demand continues to rise across industries.
Timeline and ecosystem ambitions
Rapidus aims to begin trial production of the 1.4nm process around 2029, while its 2nm node is expected to reach mass production by 2028, which indicates steady progress as the company builds out its manufacturing capabilities with support from partners such as IBM and Canon.
Japan’s broader goal involves building an end-to-end semiconductor ecosystem using local suppliers and technology, and while yield rates and production scale remain open questions, the country continues to invest heavily to position itself alongside global leaders in advanced chip development.
