As guests of Equinix’s inaugural AI summit, analysts Bob Laliberte and Savannah Peterson had front row seats to a slew of Equinix updates. Their diverse lineup of leaders made a strong case that it is innovating to enable distributed AI workloads on a global scale. Across strategy announcements, regional updates, and product innovation announcements, the company reinforced a consistent message: organizations can leverage Equinix to accelerate the adoption of AI workloads for training and, more importantly, inferencing, which supports the scale, complexity, and sovereignty requirements of global enterprises. In one striking example, 80% of Singapore’s economy flows through Equinix.
Three strategic tenets, Build Bolder, Solve Smarter, and Serve Better, were presented as the guiding framework for how Equinix is preparing to help enterprises transition from experimentation to scaled adoption of AI. These principles came to life not only through new offerings like Fabric Intelligence and the AI Solutions Lab, but also through concrete examples of expansion and customer success across the Americas, EMEA, and APAC. They’re not just building solutions, they’re teaching customers how best to achieve impact and value through their AI efforts.
Here is our on-site AnalystANGLE with real-time reactions from the Equinix Distributed AI event with Savannah Peterson and Bob Laliberte.
Building Bolder: Scaling Infrastructure for AI
The rapid rise of AI workloads has created unprecedented demands on data center design. Five years ago, most enterprise deployments ran under 5 kW per cabinet. Today, Equinix is standardizing designs at 12 kW and preparing for densities exceeding 100 kW per cabinet to support GPU-intensive training and inference.
To meet this demand, the company is embedding liquid cooling as a baseline across new builds, doubling overall capacity over the next five years, and pursuing a power-first strategy to secure reliable energy supplies. This includes expanding use of fuel cells, renewable power purchase agreements (PPAs), and longer-term nuclear partnerships.
Regional examples highlight how “Build Bolder” is translating into tangible growth:
- In the Americas, Arquelle Shaw highlighted that more than 4,000 cabinets were added in the first half of 2025, with another 8,000 expected by year-end. Flagship builds in Ashburn and Chicago will each deliver 1,500 cabinets later this year. In LATAM, Rio de Janeiro (RJ3) and Monterrey (MO2) mark critical expansions in fast-growing markets.
- In EMEA, Equinix plans to bring 8,000 cabinets online across priority metros by mid-2026, focusing on tangible delivery rather than what Bruce Owen referred to as “bragawatts” as of announced capacity.
- In APAC, Cyrus Adaggra stated the company is aggressively expanding into new markets including the Philippines, Indonesia, Malaysia, and India, while continuing to scale in Singapore, where it was the only retail operator awarded new capacity under the government’s data center moratorium framework.
These investments are designed to enable enterprises to build at scale, with the flexibility and sustainability necessary for long-term AI adoption.
Solving Smarter: Making AI Infrastructure Adaptive
While scale is essential, Equinix emphasized that success with AI also requires simplifying how infrastructure is consumed. Enterprises face complex challenges: data sovereignty restrictions, multi-provider model sprawl, and the need for <10 millisecond inference latency. It is worth noting that Equinix data centers are within 10 milliseconds of 90% of the global population. Traditional networks and static interconnection models are not sufficient.
To address this, Equinix announced the evolution of Fabric into an AI-aware, intelligent interconnection platform. The key innovation is Fabric Intelligence, which adds:
- Discovery of services and partners with one-click activation.
- Real-time visibility and AI-driven recommendations for performance, security, and cost.
- Adaptive orchestration where interconnections learn and adjust dynamically as workloads shift.
Supporting this vision are four foundational building blocks:
- Network: A global backbone spanning 270+ IBXs in 76 metros, scaling from today’s 10–50 Gbps links to 400 Gbps.
- Ecosystem: Direct access to thousands of partners and over 225 cloud on-ramps. A co-engineered AI Infrastructure Blueprint with Zayo reduces risk and accelerates multi-site deployments.
- Governance: Programmable, policy-centric networking to address compliance and sovereignty across providers and borders.
- Intelligence: AI-driven orchestration that makes interconnection adaptive, transforming it from static plumbing into a dynamic control plane.
Customers are already seeing these benefits in practice. Uber utilized Equinix Fabric to migrate 6.5 petabytes of data seamlessly between hyperscalers, thereby avoiding downtime and disruption. Hyundai Motor Group deployed its H-Cloud platform across Equinix sites in Seoul, Los Angeles, and Frankfurt, enhancing responsiveness for 10 million connected vehicles worldwide.
Serving Better: Partnering for Long-Term Success
Equinix also underscored the importance of expertise, neutrality, and operational excellence. Enterprises do not simply need infrastructure; they require trusted guidance and access to a comprehensive ecosystem to mitigate the risks associated with AI deployments.
Key elements of this commitment include:
- AI Solutions Lab: Live today in 20 locations across 10 countries, this environment allows enterprises to validate AI models, platforms, and hardware under real-world conditions before scaling into production.
- Neutral interconnection ecosystem: With nearly half a million interconnections already in place, Equinix gives customers choice and flexibility in selecting partners, meeting sovereignty requirements, and avoiding vendor lock-in.
- Operational assurance: More than 5,000 operations professionals across 36 countries support global reliability standards, ensuring customers’ mission-critical workloads remain secure and available.
Serving better also means engaging governments and communities. In the Americas, Equinix is actively collaborating with utilities and local governments to enhance grid stability while reducing time-to-power. In EMEA, the company is helping customers navigate complex regulatory requirements for data, digital, and AI sovereignty. And in APAC, partnerships with regulators and universities are driving sustainability innovation, including world-leading PUE targets in Singapore.
Our ANGLE
Few companies have the 360 view that Equinix does. They’re established, with 27 years of experience in managing both geopolitical and technical complexity. They’re known as the best-in-breed silent achievers of the data center world. Equinix’s announcements highlight a clear inflection point. As AI moves from pilots to production, the underlying infrastructure must evolve just as rapidly. Enterprises are facing new pressures: unprecedented compute densities, stringent sovereignty rules, and the need for real-time orchestration across multi-cloud, multi-region environments.
By combining scale (Build Bolder), simplification and intelligence (Solve Smarter), and ecosystem expertise (Serve Better), Equinix is positioning itself as more than a colocation provider, it is aiming to become the neutral backbone for distributed AI.
The practical takeaway is that AI adoption will increasingly hinge not only on the performance of models, but on the resilience, adaptability, and neutrality of the infrastructure underneath them. For enterprises, that means selecting partners who can deliver across the full lifecycle: designing for scale, orchestrating intelligently, and providing trusted operational support in a regulated, multi-cloud world. The exchange of ideas, best practices, and thoughtful compliance has never been more imperative.
Equinix believes it can be a strategic partner that can accelerate distributed AI initiatives on a global scale. We look forward to hearing more customer stories around the world as they surface.
For more information on Equinix solutions, please visit their website.
