In December 2025, Fujitsu and Scaleway announced a strategic collaboration to build a CPU-based AI inference option for European enterprises, Fujitsu Global reports. The MoU, signed November 27th, pairs Fujitsu's high-performance, energy-efficient Arm-based MONAKA CPU platform with Scaleway's European cloud infrastructure. The stated goals are data sovereignty, lower power consumption, and reduced supply chain risk.

This is a direct challenge to GPU-centric inference configurations. European enterprises face real pressure on both cost and compliance. A CPU-based inference path gives buyers a workload-matched alternative, with TCO as the selection lever rather than raw throughput. Scaleway is positioned as the distribution channel into that buyer base.

Seven months in, watch whether joint testing has produced workload benchmarks that hold up against GPU economics. The operating consequence is clear: inference is becoming a procurement decision, not just an engineering one. European cloud buyers now have a named CPU option to evaluate alongside existing GPU configurations.