Fujitsu Ltd announced it will begin global sales in November of "Monaka," a central processing unit (CPU) designed specifically for artificial intelligence (AI) applications and manufactured in Japan. The move represents the company’s effort to strengthen its position in the AI chip market, which has so far been dominated by overseas players.

Monaka will be integrated into domestically produced servers offered in Japan and Europe, regions where demand for secure, reliable, Japanese-made technology is growing. This comes amid a global push, particularly from the United States and China, to secure AI-related technologies amid expanding military applications of AI.

Currently, AI processing is largely powered by graphics processing units (GPUs), with Nvidia Corp, a U.S. semiconductor giant, holding a dominant market share. However, GPUs are known for their high electricity consumption during AI operations. Fujitsu positions Monaka as a more energy-efficient alternative, optimized for AI inference—the process by which AI systems generate outputs based on learned data.

Monaka leverages technology developed for Japan’s Fugaku supercomputer, reportedly delivering twice the processing performance of competing CPUs. Fujitsu claims the chip can halve both the number of servers needed and the electricity consumption required for equivalent AI workloads.

While production of Monaka will initially be outsourced to Taiwan Semiconductor Manufacturing Co. (TSMC), Fujitsu is also exploring partnerships with Japanese chipmaker Rapidus Corp for future manufacturing. The company aims to target cloud service providers and other customers across the United States and Asia as part of its broader market expansion strategy.