From Megawatts to Gigawatts: Building the AI Factories | Lenovo x NVIDIA | RAISE Summit 2026

Strategic Context and Partnership Dynamics

• John (Lenovo) leads the AI cloud and scale computing business, overseeing go-to-market and product strategy.
• David (NVIDIA) heads the Europe, Middle East, and Africa regions, having spent nearly 10 years with the company.
• Market Scale: Global data center capacity is projected to double within the next 3–5 years.
• AI Specialization: Approximately 20% of new gigawatt-scale data centers currently being built are dedicated exclusively to AI.
• Partnership Scope: The Lenovo-NVIDIA collaboration involves daily engineering co-design and co-engineering rather than periodic quarterly meetings or press briefs.
• Complexity Metric: A standard 10,000-GPU cluster requires the management of approximately 3 million cables.

Definition and Economics of "AI Factories"

• Definition: An "AI factory" is a data center optimized across five layers: land, power, shell, compute, and models/applications, functioning as a revenue stream rather than a traditional cost center.
• Primary Economic Driver: The key ROI metric for customers is "time to token," measuring the speed from concept to deployment and inference.
• Cost Optimization: Success depends on optimizing the "cost of generating that token" and throughput for large-scale inference.
• Operational Value: Without power access, AI operations are impossible; thus, securing energy is the fundamental constraint on execution.
• Investment Scale: Deploying gigawatt-scale capacity involves capital expenditures ranging from $1 billion to multi-billion dollar figures.

Execution Framework: The Four-Stage Deployment

• Stage 1 (Concept): Architects optimize data center layout, network topology, and storage to account for thermal, cooling, and power constraints.
• Stage 2 (Build): Lenovo leverages global manufacturing sites to convert component parts into working solutions near deployment locations.
• Stage 3 (Deploy): On-site teams manage physical installation, including liquid plumbing, wiring, and rack integration to ensure rapid startup.
• Stage 4 (Optimize): Post-deployment, NVIDIA NDIS services optimize models, applications, and cluster orchestration to maximize token output and economic value.

European Ecosystem Impact and Outlook

• Current Status: Europe previously lagged 12–18 months behind in compute access, a gap now actively closing through new infrastructure investments.
• Sovereign Deployments: Significant growth in sovereign AI deployments is occurring in Germany, France, and the UK to ensure data operates within national borders.
• Geographic Expansion: New data centers are being established in Northern Europe (Norway, Iceland) due to favorable energy conditions, alongside retrofitting facilities in France.
• Key Partners: N-Scale and N-Sisters are highlighted as major European partners deploying these gigawatt-scale solutions.
• Strategic Advantages for Europe:
  • Domain Data: European industries possess vast specific knowledge (healthcare, manufacturing) ideal for post-training and domain-specific AI agents.
  • Physical AI: Europe's existing robotics and manufacturing ecosystems provide a unique foundation for the next frontier of physical AI and humanoid robotics.
• Supply Chain Constraint: There is currently a 12-month waiting period for GPU access in Europe, identified by David as a critical constraint on the regional economy.
• Regulatory Bottlenecks: Uncontrollable delays regarding land permitting and power access remain the primary risk factors for deployment speed.

Lenovo's Differentiation and Future Outlook

• Single Point of Contact: Lenovo differentiates by in-house integrating concept, architecture, manufacturing, and services to control quality, security, and speed.
• Myth Correction: A prevalent misconception is that deploying AI factories is as simple as purchasing servers; in reality, it requires complex integration of liquid cooling, networking, and thermal management.
• Forward-Looking Statement: David emphasizes that Europe must "believe and act" to transform its economy, leveraging AI to turn major enterprises into true AI companies.
• Supply Chain Resilience: While Lenovo's expertise in managing complex supply chains aids stability, the sheer complexity and external factors (power/permitting) continue to present execution risks.