Kawasaki, Japan – September 16, 2026 -- Fujitsu Limited will begin global sales of its next-generation FUJITSU-MONAKA CPU in November 2026, targeting cloud providers, data center operators and server vendors with a processor designed to double AI inference throughput while halving server count and power consumption for equivalent workloads.
Fujitsu pairs the 2nm chip with a domestically manufactured server platform
Alongside the standalone CPU, Fujitsu will launch the Fujitsu MONAKA Server, built entirely in Japan and equipped with the FUJITSU-MONAKA processor. The server will roll out from November 2026 to data center operators, enterprises, and academic and HPC institutions in Japan and Europe, positioned as sovereign AI infrastructure that strengthens supply chain traceability.
The CPU hits 3.8GHz and 8800MT/s memory transfer speed
FUJITSU-MONAKA operates at a maximum frequency of 3.8GHz with a memory transfer speed of 8800MT/s, figures Fujitsu says translate into world-class computational and data-supply performance versus competing CPUs.
Air-cooled design addresses data center power and space constraints
Fujitsu says the CPU-server combination can independently build an AI inference platform, with a space-saving, power-efficient design that allows deployment in air-cooled data centers -- expanding customers' capacity to scale AI infrastructure without added cooling infrastructure. The server also supports future scalability through resource pooling technology and draws on Fujitsu's existing cooling technologies to cut cooling energy consumption.
Development was subsidized by Japan's NEDO agency
The underlying technology applied to FUJITSU-MONAKA stems from a project subsidized by Japan's New Energy and Industrial Technology Development Organization (NEDO), reinforcing the product's domestic development and manufacturing credentials.