Giga Computing Powers Scaleway's New AMD Instinct GPU Services to Accelerate AI Workloads

Powered by an Integrated AI Infrastructure Ecosystem Including Giga Computing, Bull, VAST Data, and Motul

October 13, 2026 - Giga Computing, a leader in advanced cooling servers and infrastructure solutions, today announced its collaboration with Scaleway on the launch of AI and cloud platforms using AMD Instinct™ GPUs, unveiled at ai-PULSE in Paris. Giga Computing provides the servers at the heart of the compute infrastructure, which is part of an integrated ecosystem spanning compute, storage, and liquid cooling. Scaleway’s sovereign cloud services give organizations the capacity to run sensitive and critical workloads.

⁠

Expanding Europe's AI Compute Capacity

Scaleway is building large-scale AI infrastructure in Europe, giving enterprises, research institutions, and AI developers access to high-performance cloud services in EU data centers. The AMD Instinct™ MI355X GPU adds a new accelerator option to Scaleway's AI cloud portfolio, letting customers train, fine-tune, and serve large models on an open software stack while keeping data and workloads under EU jurisdiction. The deployment also supports the Cloud Sovereignty Framework, strengthening Europe's ability to develop and run AI on infrastructure it controls.

⁠

Giga Computing at the Core of the Compute Infrastructure

Giga Computing provides the compute infrastructure that enables Scaleway's AMD Instinct-based AI & cloud services. The multi-rack deployment is built on GIGABYTE G4L3-ZX1-LAT4 DLC servers, each equipped with eight AMD Instinct MI355X GPUs and two AMD EPYC™ 9005 CPUs delivering the memory capacity and bandwidth needed for large model training and high-throughput inference. Key elements of the platform include:

• GIGABYTE GPU Servers: 4U high-density systems engineered for sustained AI and HPC workloads in cloud-scale environments.
• AMD Instinct GPU Platform: The MI355X with 288GB HBM3E per GPU, supported by the open AMD ROCm™ software ecosystem.
• Liquid-Cooling Readiness: Server designs built for direct liquid cooling, supporting higher rack power density and stable thermal performance under full GPU load.

⁠

Built for AI at Scale

Delivering AI infrastructure at scale requires more than GPU compute alone. Scaleway has brought together a partner ecosystem that spans compute, storage, and cooling: Giga Computing supplies the GPU servers, rack configuration, power, and networking that sit at the core of the infrastructure, Bull handles cluster testing and deployment, VAST Data delivers the data platform, and Motul provides closed-loop direct liquid cooling. The multi-vendor approach lets Scaleway bring a validated, production-ready AI platform to market faster than it could by building each layer in-house.

⁠

Giga Computing brings AI server expertise for large-scale deployments, backed by GIGABYTE’s manufacturing and validation. By integrating high-density GPU platforms with the management software and liquid cooling that modern data centers require, Giga Computing is helping Scaleway grow GPU capacity in step with customer demand.

⁠

“We are pleased to have Giga Computing alongside us as we expand our AI infrastructure. Their experience with high-density, liquid-cooled servers is particularly valuable for a deployment of this scale. For us, the goal is clear: give customers the compute capacity they need, with an infrastructure we can operate efficiently and grow with their ambitions,” said Albane Bruyas, Chief Operating Officer at Scaleway.

⁠

"Scaleway's AMD Instinct GPU service demonstrates the importance of bringing together high-performance compute, efficient cooling and a robust infrastructure ecosystem to meet the demands of sovereign AI at scale," said Daniel Hou, General Manager at Giga Computing. "We are proud that GIGABYTE servers form the compute foundation, and we look forward to supporting Scaleway as it expands AI compute capacity for its customers across Europe."

⁠

Giga Computing will continue to collaborate with cloud service providers and technology partners to deliver liquid-cooled, high-density AI infrastructure, helping customers move from deployment to production faster.