The October 19th, Afternoon Workshops: 1 PM - 5 PM

Intelligence Factories: Global Hubs of Compute & Capacity

Behind every headline about a new Intelligence Factory is an engineering problem most of the industry still underestimates. Power density measured in gigawatts. Cooling systems built for heat loads that didn't exist five years ago. Chip architectures, networking topology, and redundancy planning operating at a scale closer to a power plant than a traditional data center. The old timelines no longer apply. Elon Musk's compressed build schedules and NVIDIA's acceleration of chip supply have already rewritten what "fast" means in this industry — facilities that once took years are now measured in months. What used to be considered impossible is now considered a Tuesday. The companies solving these engineering problems first aren't just building infrastructure. They're setting the technical standard the entire industry will be forced to follow. This panel goes inside the build: the engineers, architects and operators actually doing the work — and rewriting the clock.

Much of the controversy around Data Center expansion misses the point: these are no longer storage-and-processing facilities. They are Intelligence Factories — purpose-built systems integrating advanced compute, semiconductors, energy, networking, cooling and software to produce intelligence at industrial scale. That distinction changes the conversation entirely. The competition for compute capacity has become a competition among states, regions and nations for technological leadership, scientific advancement, industrial development and economic power. Building it requires unprecedented coordination among hyperscalers, energy producers, semiconductor companies, infrastructure developers, government and global capital. The Intelligence Factory isn't the next generation of the Data Center. It is one of the defining industrial institutions of the Intelligence Economy — and the jurisdictions that understand this first will own the decade ahead

No single company — however large — can build the Intelligence Economy alone. The scale of capital required for compute, energy and infrastructure has forced a new kind of alliance: hyperscalers partnering with sovereign wealth funds, state governments competing to underwrite gigawatt-class campuses, banks structuring project finance at a scale once reserved for pipelines and power plants, real estate investment groups racing to secure land and power interconnects, and venture funds staking early positions before capacity is spoken for. These aren't traditional vendor relationships. They are structural partnerships redefining who owns, who operates and who profits from the infrastructure of intelligence itself — negotiated faster, and across more players, than markets have seen in a generation. The winners of the next decade won't be the companies with the best technology alone. They will be the ones who assembled the right partnerships — states, banks, real estate, capital and technology moving as one coordinated bet on the future.

Compute has a physics problem: heat, power and land are finite on Earth, and the Intelligence Economy is already running into all three. The next frontier isn't hypothetical — it's orbital. Solar-powered compute constellations, off-Earth data processing, and a satellite-based intelligence internet are moving from research papers to real capital allocation, with Wall Street and Silicon Valley both beginning to underwrite the pioneers. This is not a distant moon-shot conversation. It is a near-term strategic question: which companies, which nations, and which investors position themselves early in an economy where compute capacity may no longer be bound by a single planet's energy grid.

The Earth/Off-Earth Economy asks the question every serious operator in this space is already asking privately: when compute goes off-world, who gets there first — and who profits from having arrived early.

The gigawatt campuses being built across Texas, Ohio and the frontier states exist for one purpose: to make the impossible computable. But raw compute alone does not produce discovery, art, or meaning — it produces capacity. The defining question of the Intelligence Economy is not how much compute we can build, but what happens when that capacity meets human ambition, imagination and will. This is where the Intelligence Factory stops being an industrial story and becomes a creative one. The scientists solving fusion, the engineers modeling protein folding, the writers and architects using super intelligent tools to compress decades of work into months — all are drawing from the same well of compute, aimed at entirely different frontiers. Scientific ambition and human creativity are converging inside the same infrastructure. The Intelligence Factory doesn't replace the impossible dream — it becomes the engine that finally makes it reachable.

The assumption is that Intelligence Factories compete on technology — better chips, better cooling, better architecture. The reality is different. The real moat is capital deployed at a scale most institutions have never attempted, political will sustained long enough to secure permitting and power allocation, and the conviction to commit billions before the outcome is proven. Technology can be licensed, copied, or purchased. Commitment cannot. The regions and nations pulling ahead aren't necessarily the most technically sophisticated — they're the ones whose governments, capital markets and institutions aligned fastest and stayed aligned longest. The cost of standing still isn't technical obsolescence. It's political and financial paralysis while others move. This panel examines what actual commitment looks like at gigawatt scale — and why hesitation, not technology, is the real threat to falling behind.