IBM Says AI Hasn’t Killed Mainframes After 42% Sales Plunge
IBM’s mainframe franchise underpins mission-critical systems at banks, airlines and governments, while sales of IBM Z machines typically pull through lucrative transaction-processing software. That makes the business a key test of whether surging AI infrastructure spending is merely changing the timing of enterprise purchases or structurally eroding demand for legacy computing. Customers’ late-quarter shift toward servers, storage and memory squeezed budgets for IBM’s higher-margin hardware and software stack, sharpening investor concern over the company’s growth model.
IBM on July 22 reported second-quarter revenue of $17.162 billion for the period ended June 30, up 1% from a year earlier, and adjusted earnings of $2.93 a share. IBM Z revenue plunged 42%, dragging infrastructure sales down 7%. The shares had tumbled about 25% on July 14 after the company preannounced the miss. CEO Arvind Krishna and CFO James Kavanaugh said delayed deals and budget shifts were temporary and rejected the view that AI was killing mainframes. IBM lowered its 2026 constant-currency revenue-growth outlook to 4%–5%.
All Coverage
1 original reportsThe Backstory
The history behind this eventIBM Misses Forecasts as Customers Rush to Buy AI Servers and Memory
The infrastructure arms race fueled by generative AI is drawing large portions of corporate IT budgets toward scarce AI servers and high-bandwidth memory, crowding out spending on traditional software and mainframes. With businesses spanning enterprise software, consulting services and mainframes, IBM is viewed as a key gauge of how this “AI crowding-out effect” is affecting established IT companies. Its results also have a direct bearing on market confidence in the broader software sector.
In a rare move, IBM released preliminary second-quarter 2026 results ahead of its formal earnings date and acknowledged that both revenue and earnings would fall short of market expectations. Its shares plunged by about 24% to 25% at one point in premarket trading, dragging software stocks broadly lower and prompting the market to describe the sector as an “AI casualty.” Chief Executive Arvind Krishna said corporate customers had shifted substantial portions of their budgets to secure servers and memory in tight supply, crowding out planned software and mainframe contracts and causing the quarterly shortfall.
Subscribe to Mark Radar Weekly
Every Friday, the week's strongest signals in your inbox. Unsubscribe anytime.