The beginning of July 2026 marks a major strategic shift for the technological ecosystem of the Iberian Peninsula. Within a 48-hour span, two groundbreaking announcements have emerged: Microsoft’s confirmation of a $10 billion investment to establish a large-scale AI and cloud data center campus in Sines, Portugal, and the official launch of "Amália", the region's first open-source large language model tailored specifically for Iberian linguistic contexts and European regulatory frameworks.
The convergence of these two projects forms what is now known as the "Iberian AI Corridor." For European SMEs, this infrastructure offers an unprecedented opportunity to address three of the biggest roadblocks to corporate AI adoption: latency, strict GDPR compliance, and reliance on foreign technology providers.
We analyze the technical specifications of this new infrastructure, the capabilities of the Amália model, and how local businesses can position themselves to leverage this regional advantage.
1. The Sines Gigafactory: Localized Supercomputing for AI
Microsoft's campus in Sines is not a conventional web hosting facility. It is engineered specifically as an "AI computing gigafactory" built to support the next wave of intelligent agents and abstract reasoning models.
Capacity and Hardware
The facility, developed in collaboration with Start Campus and Nscale, aims to reach a computing capacity of 1.2 Gigawatts (GW) by 2030. In its initial phase, the center will house 12,600 NVIDIA Blackwell Ultra GPUs, with plans to expand the second building to include over 66,000 NVIDIA Rubin GPUs—representing the most advanced hardware architecture available.
Sustainability and Seawater Cooling
Leveraging its coastal location, the campus utilizes seawater cooling systems and runs entirely on local renewable energy. This enables Sines to offer highly stable, energy-efficient inference and processing costs compared to traditional data centers located in Northern Europe or the US.
2. "Amália": The Open-Source Foundation for Iberian Digital Sovereignty
Concurrently with Microsoft's physical infrastructure deployment, Portugal’s public research consortium (backed by €5.5 million from the country's Recovery and Resilience Plan) has officially launched Amália (Automatic Multimodal Language Assistant with Artificial Intelligence).
This is the first large language model (LLM) designed from scratch to align with the linguistic, cultural, and regulatory realities of the Iberian Peninsula.
- Native Legal Compliance: Unlike US models trained under Anglo-American fair use doctrines, Amália has been trained in strict compliance with European copyright laws and the simplified Digital Omnibus on AI regulatory framework enacted by the European Parliament.
- Institutional and Corporate Integration: Developed as a foundation platform, it allows local businesses and public administrations to deploy internal agentic solutions safely, ensuring the model understands the specific administrative, tax, and legal frameworks of our region.
🔒 Worried about data privacy when routing corporate assets to foreign APIs?
The deployment of the Sines data center and the launch of sovereign models like Amália prove that you can integrate advanced AI while maintaining 100% digital sovereignty. At IA4PYMES, we audit your architecture, build GDPR-compliant integrations, and configure your systems to connect with the new low-latency Iberian corridor.
Book your 60-minute technical consultation here (100% refundable if we don't validate project feasibility in 15 minutes, or fully credited against final development costs on hire).
3. Key Competitive Advantages for Tech SMEs
The launch of the Iberian AI Corridor improves both the operational and financial viability of AI integrations:
Mitigating International Data Transfer Risks
Under current EU regulations, sending medical records, financial histories, or customer databases to APIs with US-based servers presents a continuous GDPR compliance risk. Running agentic workflows on the Sines node guarantees that sensitive data never leaves the European Union, protecting SMEs from legal audits and potential compliance penalties.
Ultra-Low Latency for Real-Time Agents
Automated voice systems and real-time agents operating in industrial or customer support settings require immediate response times. The physical proximity of the Sines data center reduces network transit times to minimal levels, enabling fluid, real-time interactions that were previously hindered by transatlantic routing latency.
4. Strategic Roadmap: Preparing for Localized AI
Businesses looking to lead in this new environment must adapt their technical architectures immediately:
Design Iberian and Hybrid Cloud Architectures
If you leverage Microsoft Azure for your AI pipelines, configure your resource groups and cognitive endpoints to route requests to Iberian nodes as soon as they become commercially available. This ensures immediate performance and latency improvements without additional development costs.
Evaluate Regional Open-Source Models
Begin testing open-source regional models like Amália or Qwen 3.6 Coder for local document processing and business intelligence. The ability to fine-tune these models on-premise using proprietary corporate data offers a level of security and business continuity that closed commercial APIs cannot match.
Conclusion
The Iberian AI Corridor represents our region's response to global technological dependence. By combining Microsoft's massive supercomputing scale in Sines with the development of sovereign open-source models like Amália, the Iberian Peninsula is establishing itself as an ideal environment for secure, cost-effective, and high-performance business AI. SMEs that adapt their infrastructures today to leverage this low-latency node will gain a direct competitive advantage in operational cost and regulatory compliance.