Where AI Leaders Meet: Rafay’s Infrastructure Leadership Summit in Barcelona

Capacity Is the Starting Line, Not the Destination

The largest infrastructure build-out of our era is underway, and it is about to separate the operators who simply own compute from the ones who build lasting businesses on top of it. Most operators are still at the starting line, focused on racking hardware and bringing capacity online. But the destination is not more GPUs. It is the journey from GPU to token, where raw compute becomes governed AI services that customers consume and pay for. AI infrastructure is entering that defining moment right now.

Building a neocloud is a journey, not a purchase. It begins with standing up the investment: racking GPUs, wiring high-performance networking and storage, and bringing bare metal online as a service customers can reach on their own. The next step turns that foundation into something consumable, adding self-service access, multi-tenancy, governance, and the orchestration that lets a small platform team serve many customers at once. Each layer adds value the one below it could not.

The destination is monetization. Bare metal as a service (BMaaS) becomes GPU-as-a-service, then packaged SKUs, metered usage, and billing that converts consumption into revenue. At the top of the journey sits inference, where GPUs become token-metered AI services that customers pay for by the token. This is the path from BMaaS to tokens, and it is where durable margins and lasting customer relationships are built. It is also why the world's leading neoclouds run on Rafay.

The AI Infrastructure Leadership Summit brings those operators into one room. Hosted by Rafay from 8th to 10th September 2026, it is where teams learn from peers who have walked the path, sharing what works and where the real challenges lie on the way from GPUs to tokens.

Who Is in the Room

It is two days with the people actually building the next generation of AI businesses in one place: founders and CEOs who have taken an idea to real revenue, the executives shaping where the ecosystem goes next, and the operators serving enterprises, developers, and entire nations. The value is as much in the hallway conversations and the roundtables as in the sessions themselves. Leaders from across the stack are on the program, including NVIDIA, Lenovo, WWT, DDN, and Cisco, alongside global integrators like Accenture who connect this infrastructure to the enterprises where AI demand ultimately lands, and the founders and operators building AI businesses today.

Day 1 is about the economics

What it takes to build an AI infrastructure business that is profitable, not just busy. Where margin actually comes from. How the best operators moved from selling GPU time to selling outcomes, and what that decision did to their economics. The people on stage have already done it, and they are sharing the specifics. DDN, for one, shows how the data layer quietly sets the ceiling on utilization and margin.

Day 2 is about what comes next

The move up the stack from GPU hours to inference, tokens, and outcomes. What it means when an AI cloud grows into a full platform business, a path NVIDIA has walked at global scale. Lenovo makes the case for private AI and the shift from maximizing tokens to maximizing outcomes. And the leaders serving regulated and sovereign markets show how they are turning purpose into a durable advantage. It closes on the decisions that separate an infrastructure project from a business that lasts.

What the Summit Offers Each Audience

For neocloud and GPU service providers, the summit is the clearest answer to margin compression, showing how to move from selling hours to selling governed, token-metered AI. For telcos, it is a path to turning regional infrastructure and edge presence into consumable AI services with the governance carriers require. For sovereign AI operators, it is a room full of people proving that regional ambition can become a real business, not just national capacity. And for enterprise AI leaders, it is a map of what it takes to serve AI at scale with the trust and control the business demands.

The AI infrastructure market is maturing fast, and capacity alone is no longer the advantage it once was. The operators building lasting businesses are the ones moving past raw compute into governed, monetizable AI services.

Reserve your seat at Rafay’s AI Infrastructure Leadership Summit, 8th to 10th September 2026. We look forward to seeing you there.