Fujitsu Unveils MONAKA 144-Core AI Server CPU for November Release
Fujitsu's MONAKA CPU, a 144-core Armv9 processor, will launch in November, featuring advanced 2nm technology and high performance for AI tasks.
Safety level: informational · Level: beginner
Fujitsu Introduces the MONAKA 144-Core Armv9 Server CPU
Fujitsu has announced the launch of its new MONAKA CPU, which boasts an impressive 144 cores based on the Armv9 architecture. This server processor is manufactured using an innovative 2 nm 3D-stacked process, which integrates a 2 nm core alongside a 5 nm cache and I/O layer. This design aims to provide enhanced power efficiency and performance, particularly for artificial intelligence (AI) inference tasks.
The MONAKA CPU is capable of reaching a maximum clock speed of 3.8 GHz and supports memory speeds of up to 8,800 MT/s. It is equipped with hardware acceleration specifically designed for matrix operations, and it utilizes SVE2 vector instructions. This combination allows the MONAKA to achieve approximately double the AI inference throughput when compared to other processors in its class.
Fujitsu will offer the MONAKA CPU in a 1U air-cooled rackmount server configuration, where all 144 cores operate at a frequency of 2.1 GHz. Additionally, there is a liquid-cooled version that can boost performance up to 2.9 GHz. Each socket of the server supports twelve RDIMM slots and is compatible with PCIe Gen6, ensuring high-speed connectivity for various applications.
The MONAKA CPU is set to be available starting in November, with plans for broader distribution across Europe anticipated by 2027. This launch signifies a significant leap forward in server hardware optimized for AI applications, presenting a powerful combination of processing capabilities and memory efficiency in a compact and energy-conscious design.
Implications for Buyers in India
For those considering a purchase in India, the MONAKA CPU represents a noteworthy option for enterprises focused on AI workloads. Its advanced specifications make it suitable for data centers and businesses looking to enhance their computational capabilities. However, potential buyers should evaluate their specific needs, as the high-performance features may be more relevant for organizations heavily invested in AI rather than for general computing tasks.
Still stuck?
If anything here does not match what you are seeing, stop rather than force it. Contact us · Book a repair
Last reviewed 2026-11-10.