IBM Warns AI Infrastructure Shift Impacts Software Deals, Shares Plunge

IBM shares plunged 23% after the company warned that customers are shifting spending to secure supply-constrained AI infrastructure, delaying software and consulting deals. CEO Arvind Krishna said the capex reprioritization emerged late in the quarter, with distributed infrastructure revenue growing 37% while overall infrastructure fell 7%. The shift suggests enterprises are prioritizing hardware for AI buildout before expanding software spending.

IBM Warns AI Infrastructure Shift Impacts Software Deals, Shares Plunge IBM said customers shifted spending to secure supply-constrained infrastructure, delaying software deals and driving shares down 23%. IBM shares plunged about 23% on Tuesday after the company warned second-quarter revenue and earnings would miss Wall Street expectations. The company said customers are redirecting capital to servers, storage, and memory while software and consulting deals slipped. IBM expects second quarter revenue of $17.2 billion, up 1% year over year, but below analysts’ expectations. Software revenue rose 5%, consulting revenue was flat and infrastructure revenue fell 7%. Adjusted earnings per share are expected to reach $2.93. Chief Executive Officer Arvind Krishna said the customer capex change emerged late in the quarter. “In the last few weeks of June, we saw clients shift their quarterly capex spend toward servers, storage, and memory purchases to secure supply-constrained infrastructure ahead of expected price increases,” Krishna wrote in a letter to investors https://newsroom.ibm.com/2026-07-14-Arvind-Krishnas-Letter-to-IBM-Investors . “While we anticipated some supply chain related impact… we did not anticipate the magnitude of the capex reprioritization.” The CEO also cited industry-wide cybersecurity issues that distracted customers during the quarter and acknowledged IBM’s execution. “We did not adapt and move quickly enough, and numerous large deals failed to close on the timelines we expected, driving the majority of our shortfall,” Krishna wrote. The comments suggest customers did not abandon AI spending. Instead, they redirected capital toward hardware purchases as they sought to secure constrained infrastructure before anticipated price increases. Distributed Infrastructure Grows While IBM’s infrastructure segment declined 7%, its distributed infrastructure business posted its strongest quarter on record. Krishna said distributed infrastructure revenue grew 37%, driven by power systems and storage, with a $500 million backlog at quarter’s end. The contrast suggests customers continued investing in cloud and edge environments even as IBM’s z17 mainframe /servers/ibm-brings-z17-mainframes-to-standard-racks cycle and the associated software stack underperformed expectations. For data center operators, IBM’s update offers little evidence that enterprise investment in AI infrastructure is fading. Instead, it suggests organizations continue securing compute, storage, and supporting infrastructure before expanding spending on software and services. That sequence could sustain demand for power, high-density capacity, storage and networking infrastructure even if some enterprise software projects move more slowly. Dave McCarthy, research vice president of cloud and edge infrastructure services at IDC, told Data Center Knowledge the company is seeing the same pattern across the enterprise market. “IDC is absolutely seeing an intensive ‘infrastructure-first’ phase across the broader enterprise market,” McCarthy said. “Many organizations are aggressively routing capital expenditure to secure high-performance servers, storage, and memory. They are doing this not just to keep pace with AI demands, but to lock in supply and pricing before expected cost hikes.” Neil Osnato, founder of Persistence Analytics Group, told Data Center Knowledge IBM’s results reflect a broader sequencing issue rather than weakening enterprise interest in AI. “Many organizations are still in the AI infrastructure buildout phase,” Osnato said. “They are prioritizing servers, storage, memory, data center capacity, power availability, and physical execution before software and consulting spend fully scales.” “AI demand is real, but the spending sequence is uneven,” he said. “Hardware and infrastructure can pull forward while software and services slip.” Execution or Market Shift? Zeus Kerravala, founder and principal analyst at ZK Research, told Data Center Knowledge the infrastructure spending trend does not fully explain IBM’s results. “I’d see this as primarily an IBM execution issue playing out against a very real AI-driven infrastructure spending cycle,” Kerravala said. While customers are prioritizing investments in servers, storage and memory to support AI workloads, vendors with compelling AI and hybrid cloud strategies should still be closing software and consulting deals, he said. “That gap suggests IBM is still struggling to turn its AI story into a differentiated sales motion and business outcomes customers are willing to fund on the software and services side right now.”