The article argues that the AI-driven shortage of servers, memory, storage and networking equipment is giving hyperscale cloud providers greater control over the enterprise hardware supply chain. Because firms such as Amazon, Meta and other major cloud operators can place large, long-term orders, businesses that need new infrastructure quickly may increasingly have to rent it from those providers instead of buying it outright.

Nutanix CEO Rajiv Ramaswami said renting from a hyperscaler can be faster than waiting for conventional hardware suppliers, while Micron, SK Hynix and Seagate have secured supply agreements with major customers. Amazon says AWS typically recovers server and networking investment within three years, despite equipment lasting at least five to six years and AI capacity often being contracted for five years. Traditional suppliers including Dell, HPE, Lenovo and Supermicro can still offer more predictable bills, but with less flexibility than cloud capacity.