After shocking quarter, IBM insists that AI isn’t killing the mainframe

IBM used its quarterly earnings call this week to push back against a narrative that has rattled investors: the idea that corporate spending on artificial intelligence is starving its legacy mainframe business. Chief executive Arvind Krishna told analysts that any pullback in mainframe orders reflects a temporary reshuffling of technology budgets toward AI projects, not a permanent decline in demand for the room-sized systems that still anchor banking, insurance, and government computing worldwide.

The reassurance came just days after IBM shares tumbled following the company's own warning that results would land below what markets had been braced for. That early heads-up from Krishna and the board was itself notable: preemptively flagging a shortfall before the formal earnings release is a step the 115-year-old technology giant had not typically taken, and it primed investors for a rough set of numbers when Wednesday's report finally arrived.

Those numbers turned out largely as advertised. IBM posted $17.2 billion in quarterly revenue alongside $9.9 billion in gross profit, translating to a robust gross margin close to 58%. Net income for the period came in at $2.2 billion. On their own, those figures describe a company still generating enormous cash flow. Measured against what Wall Street had modeled, however, the results fell well short, deepening concern that IBM's core infrastructure lines are losing momentum faster than analysts had assumed.

Mainframes have long been a comparatively small but highly profitable slice of IBM's business, prized by large enterprises for handling transaction volumes and compliance workloads that many companies remain reluctant to migrate to the cloud. Because mainframe refresh cycles tend to be lumpy — clustered around new hardware generations rather than spread evenly year to year — even a modest slowdown in new orders can distort quarterly comparisons and unsettle investors who track the segment closely as a bellwether for IBM's broader health.

Krishna's explanation attempts to reframe that slowdown as a byproduct of success elsewhere rather than a sign of structural decline. His argument is that chief information officers are currently redirecting discretionary hardware and infrastructure budgets toward building out generative AI capabilities, leaving less near-term room for mainframe upgrades even among customers who are expected to eventually return to them.

Not every observer is likely to accept that framing at face value. Skeptics on Wall Street have questioned whether IBM's AI-related businesses, including its Watsonx platform and consulting arm, are growing quickly enough to offset softness in infrastructure sales, and whether describing the mainframe dip as "temporary" is easier to assert than to prove within a single earnings cycle. The unusual decision to warn investors ahead of the formal report suggests IBM itself recognized the risk of a harsher market reaction without some advance context.

Attention now shifts to whether IBM's next mainframe hardware cycle and its AI-services pipeline can validate Krishna's timeline. Investors and customers alike will be watching subsequent quarters for evidence that enterprise budgets swing back toward infrastructure spending, or whether the current dip proves more durable than the company is currently willing to concede.

Source: TechCrunch