IBM says AI spending delayed mainframe purchases after weak quarter

In short: IBM says a sharp drop in mainframe sales was caused by customers delaying purchases as AI-related hardware costs and budgets shifted.

What happened

IBM reported quarterly earnings that fell well short of Wall Street’s expectations. The company still posted $17.2 billion in revenue and $2.2 billion in net earnings, but the miss rattled investors.

A key reason was IBM’s “infrastructure” business, where mainframe hardware sales fell 42%. Mainframes are large, expensive computers used by big organizations to run critical work like payments and records (think of them as a central, heavy duty engine in a company’s data center).

IBM CEO Arvind Krishna and CFO Jim Kavanaugh said on an investor call that “tens” of customers who were expected to buy new mainframes in the quarter decided to wait. Even a small number of delays can hurt results, because a single mainframe can cost from hundreds of thousands to millions of dollars.

Krishna blamed budget pressure created by the current rush to buy AI-related equipment. He said customers faced price increases of about 15% to 30% for other data center gear and PCs, so they moved money there first instead of buying a new mainframe.

IBM also lowered its full-year growth forecast. Kavanaugh added that mainframes matter beyond hardware, because IBM says it earns about $3 in software revenue for every $1 of mainframe hardware sold.

Why it matters

This is a reminder that even when AI spending is rising, it can push other purchases into “later” rather than “never.” IBM is arguing that customers are not abandoning mainframes, they are just postponing big upgrades, like putting off a car purchase after paying for an expensive home repair.

Source: TechCrunch AI