Here's the thing about Oracle Corp (ORCL) and AI: everyone agrees the demand is there. The question is whether the company can actually build fast enough to catch it.

Gil Luria of DA Davidson laid it out pretty clearly in a recent interview. He says Oracle is in a sweet spot—or maybe a high-pressure spot—in the AI infrastructure race. It's not quite in the top tier with the hyperscalers like Amazon.com Inc (AMZN), Microsoft Corp (MSFT), and Alphabet Inc's Google (GOOGL). Those guys own the primary customer relationships. But Oracle isn't some scrappy startup either. It's the go-to overflow provider, sitting much closer to the giants than to the smaller "neo-cloud" players nipping at their heels.

What gives Oracle that heft? Its classic, cash-generating software business. That's the engine funding its massive push into building data centers for the AI boom. It's a nice position to be in: using reliable old business money to bet on the shiny new thing.

The Backlog Tells the Story

The most eye-popping number here is the backlog. Luria highlighted that Oracle's backlog stands at about $550 billion. Let's put that in perspective. He notes that Microsoft Azure's backlog is slightly over $600 billion, while Amazon Web Services is sitting at roughly $350 billion. That puts Oracle firmly in second place in this particular metric.

A $550 billion backlog isn't just a number on a spreadsheet; it's a giant pile of promised future revenue. It means customers—big AI companies—are essentially lining up and saying, "We need your computers, please build them for us." That kind of visibility is every CFO's dream. It suggests that, for now at least, demand isn't the problem.

So What's the Catch? Execution.

Of course, a backlog is just a promise until you deliver. Luria acknowledged the elephant in the server room: will all this projected demand actually materialize? There's always a risk that the AI hype cycle cools, or that companies scaling their models hit a wall.

But he argued that recent news has actually reduced that risk. He pointed specifically to Microsoft-backed OpenAI raising a staggering $122 billion. That kind of war chest, Luria says, can fund its compute needs—its need for powerful servers and data centers—for the next two to three years. And a big chunk of that spending is committed to providers like Oracle, Microsoft, and Amazon.

Add in rising revenues at OpenAI and its rival Anthropic, plus the fact that there's still a global shortage of AI chips and computing power, and the demand picture looks pretty solid. The bottleneck, then, shifts. The main challenge for Oracle is no longer finding customers who want its services; it's actually building the physical infrastructure to provide them.

Can they execute? Can they get those data centers built, powered, and online fast enough to capture the value of that $550 billion backlog before the window closes or competitors catch up? That's the multi-billion dollar question. It's a construction and logistics challenge on a monumental scale.

Investors were chewing on this very question Monday morning. Oracle shares were down 0.94% at $173.42 in premarket trading, according to market data. A tiny move, really, but a reminder that in this market, promise is priced in. The real money is made—or lost—on the follow-through.