Recommended Articles
Fundraising

Analysis: The AI Money Is Moving From Chips to Everything Around Them

By Angeled · October 24, 2026 · Fundraising
Analysis: The AI Money Is Moving From Chips to Everything Around Them
Analysis: The AI Money Is Moving From Chips to Everything Around Them

Two announcements on October 23, 2026, tell the same story from different angles. Bank of America Global Research projected that global data center spending will pass $1 trillion in 2026 and more than double to $2.1 trillion by 2028. On the same day, Oracle announced a $1 billion commitment over four years to its partner ecosystem, funding training, technical enablement, cloud migration, and go-to-market collaboration for the companies that sell and implement AI solutions.

Read together, they mark a shift in the AI investment cycle. The first phase was about chips: who could design and manufacture the processors that train and run AI models. The next phase is about everything around the chips — and it is bigger, messier, and harder to finance than the first.

What the Numbers Actually Say

Bank of America's forecast contains a detail that reframes the whole buildout. The bank estimates data center construction at about $49 million per megawatt of power capacity, with IT equipment accounting for 78% of total spend. Servers alone cost about $31.4 million per megawatt, and networking adds another $4.8 million. The remaining items — switchgear, power distribution, generators, backup power, chillers, and cooling — cost between $100,000 and $1.1 million per megawatt each.

In other words, the AI buildout is still mostly a computing purchase, not a construction project. That is good news for chipmakers and server builders, which have been among the biggest beneficiaries so far. But it also means the buildout's pace depends on the supply chains, lead times, and pricing of servers and networking gear — not just on how much concrete gets poured.

The bank also projects the data center market growing from $275 billion in 2022 to as much as $2.93 trillion by 2030, a compound annual growth rate of 37% from 2025 through 2030, driven by AI demand. AI-focused "neocloud" data centers, the bank estimates, cost about $57 million per megawatt but need less space than traditional facilities. Those figures describe an industry scaling at a pace rarely seen outside wartime mobilization — and raising the question of whether power grids, supply chains, and financing can keep up.

Why Oracle's $1 Billion Matters

Oracle's commitment looks small next to trillion-dollar forecasts, but it addresses a bottleneck the forecasts do not capture: delivery capacity. Oracle's cloud infrastructure business has grown rapidly, and the company says its partners extend its sales reach and expand its delivery capacity. A data center is only useful when somebody can migrate a customer onto it, integrate it with existing systems, and keep it running.

That is what partners do: systems integrators, independent software vendors, cloud and managed service providers, resellers, and advisory firms.

Where the Next Opportunities Are

For founders, the shift from chips to everything-around-chips opens several kinds of opportunity.

First, power and cooling. If data center spending reaches the trillions, the physical inputs — power generation and distribution, cooling systems, backup power — become large markets in their own right, even though they are a smaller share of each project's cost. Founders with expertise in energy systems, grid interconnection, or industrial cooling are suddenly selling into the fastest-growing capital market in the world.

Second, deployment and integration services. Oracle's partner investment signals that the market will pay for companies that help customers adopt AI infrastructure.

Third, efficiency. At $49 million per megawatt, every percentage point of efficiency is worth an enormous amount of money.

What Investors Should Watch

For investors, the Bank of America forecast and Oracle's commitment suggest a rotation in where value accrues. The early winners were chipmakers and server builders. The next winners may include industrial suppliers, power and cooling companies, neocloud operators, and the services firms that deploy AI systems. That rotation does not mean chip stocks are finished — servers are still the largest cost line — but it does mean the investment map is getting wider.

Investors should also watch for the risks that come with scale. Trillion-dollar forecasts assume that power, financing, and demand all keep pace.

The AI buildout began as a chip story. It is becoming an infrastructure story, an energy story, and an ecosystem story. Founders and investors who follow the money past the chips — into servers, power, cooling, and the partners who make it all work — will see the next phase of the opportunity more clearly than those still staring at the silicon.

Angeled.com connects founders with investors.

Reuters via TradingView/Q1019FM, October 23, 2026: "Data center spending to more than double to $2.1 trillion by 2028, BofA says"
PR Newswire via Morningstar, October 23, 2026: "Oracle Announces $1 Billion Investment to Accelerate Partner Growth and AI-Driven Modernization for Customers"
Dow Jones Newswires via Morningstar, October 23, 2026: "Oracle Commits $1 Billion to Help Partners Improve AI Offerings"

Read Next