Oracle just proved that in the AI gold rush, even old shovels sell at a premium. The company disclosed during its Q1 FY2027 earnings call that it renewed or resold all of its expiring GPU capacity at prices roughly 20% higher than prior contracts. The kicker: most of those GPUs are over four years old.
The numbers behind the pricing power
Oracle’s GPU utilization rate hit 97.9% during the quarter. The demand side of the equation is equally striking. Oracle booked more than $30 billion in new AI cloud contracts during Q1, pushing its remaining performance obligations to $664 billion. That figure rose $26 billion from the prior quarter and $209 billion compared to a year ago.
Oracle delivered more than 300,000 GPUs to its AI cloud customers during the quarter, tripling the output from Q4 FY2026. The company also brought an additional 850 megawatts of AI capacity online.
The client roster reads like a who’s who of the AI arms race: OpenAI, xAI, Meta, NVIDIA, and AMD.
Why old GPUs are still commanding premium prices
The AI industry faces acute shortages of new NVIDIA GPUs, creating a bottleneck that has rippled through every major cloud provider. Oracle’s strategy effectively exploits a temporary supply-demand mismatch. Customers who locked in GPU capacity years ago at lower rates are discovering that walking away from those contracts means entering a market where replacement capacity may not exist at any price. Renewing at a 20% markup suddenly looks like a bargain compared to the alternative of losing access entirely.
Oracle raises its financial outlook
Buoyed by these results, Oracle raised its full-year FY2027 revenue guidance to at least $90 billion. That represents 34% year-over-year growth. The company also lifted its non-GAAP earnings per share outlook to $8.10.