The FUJITSU-MONAKA CPU uses a 3D-stacked architecture with a 2 nm core and 5 nm cache, achieving a maximum clock speed of 3.8 GHz and memory transfer speeds of 8800 MT/s. Fujitsu claims the processor delivers twice the AI inference throughput of other CPUs while halving the power consumption and physical footprint required for equivalent workloads.

The accompanying MONAKA Server is positioned as a sovereign AI platform, offering domestic manufacturing for supply chain transparency. It supports future scalability through resource pooling and is designed to operate in air-cooled data centres, addressing the growing need for on-premises AI infrastructure without specialised cooling.