Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
Gartner’s widely cited 7.9% figure was a July 2025 forecast for worldwide IT spending that year—not a measure of spending growth in every category and not the latest outlook. Gartner projected $5.43 trillion in 2025 spending, with AI-related infrastructure helping drive the increase. Its latest forecast cited here, published July 27, 2026, calls for 14.2% growth in 2026, to $6.37 trillion. For investors and household-budget watchers, the important point is that these are changing market forecasts, not guarantees of company profits, investment returns or lower consumer costs.
What Gartner’s 7.9% forecast measured
On July 15, 2025, Gartner forecast that worldwide end-user IT spending would reach $5.43 trillion in calendar year 2025, a 7.9% increase from 2024. Its total spans data-center systems, devices, software, IT services and communications services. It is not a forecast that every category—or every company’s technology budget—would rise by 7.9%, and it is not an AI-spending figure.
Gartner attributed much of the outlook to accelerating investment in AI-related infrastructure, particularly data-center systems and AI-optimized servers. The release also described an “uncertainty pause” affecting some software and services decisions. Read the original Gartner forecast with its date attached: the percentage applies to 2025, as forecast in July 2025.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
The latest forecast is higher—and for a different year
Gartner revised its outlook as expectations for AI infrastructure, cloud services and data-center investment shifted. Its successive forecasts for 2026 were:
#1 Best Overall
| Forecast published | Year forecast | Worldwide IT spending | Forecast growth |
|---|---|---|---|
| July 15, 2025 | 2025 | $5.43 trillion | 7.9% |
| February 3, 2026 | 2026 | $6.15 trillion | 10.8% |
| April 22, 2026 | 2026 | $6.31 trillion | 13.5% |
| July 27, 2026 | 2026 | $6.37 trillion | 14.2% |
The 2026 figures are successive forecast snapshots, not reported final results. The February forecast projected data-center spending growth of 31.7%, to more than $650 billion; in April, Gartner projected data-center systems spending above $788 billion. The July release identified data-center systems and infrastructure as a service (IaaS) as leading growth segments. See Gartner’s February, April and July releases for their respective assumptions and dates.
For context, Gartner had earlier forecast 9.3% growth for 2025 in October 2024, then 9.8% in January 2025, before its July 2025 7.9% revision. That sequence illustrates why a forecast should be treated as a dated estimate that can change—not as a settled outcome.
Rank #2
What an “infrastructure revolution” means in practice
AI services require computing capacity, and the spending response reaches well beyond buying graphics processors. AI-capable infrastructure can involve:
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Clear out junk files and repair common Windows errors3Fix the driver behind crashes, sound loss and screen glitches- Accelerated servers: GPU-based systems and custom AI chips, alongside host processors and high-bandwidth memory.
- Networking and storage: Fast connections between accelerators, plus systems to feed and retain large datasets.
- Data centers: Buildings, racks and equipment designed for dense computing loads.
- Power and cooling: Electricity delivery and thermal management, which can constrain how much equipment a site can run.
- Cloud infrastructure: Providers investing in capacity and selling it as compute instances, managed platforms and inference services.
- Software and operations: Orchestration, security, monitoring and data pipelines needed to make capacity usable.
The mix matters as much as the headline total. Gartner’s July 2025 release said spending on AI-optimized servers, negligible in 2021, was expected to become roughly three times traditional-server spending by 2027. That was a forecast, not a report of an already completed shift or a prediction that conventional servers would disappear.
Similarly, AI infrastructure is a major growth driver, but it is only part of worldwide IT spending. In a separate May 19, 2026 forecast, Gartner projected worldwide AI spending of $2.59 trillion for 2026, up 47%, and said AI infrastructure would account for more than 45% of that AI-spending total. This is a different market measure from total IT spending; the two figures should not be added together as though they were separate pots of money. Gartner had published an earlier January 2026 AI forecast of $2.52 trillion, another reminder to date each estimate. See the May AI forecast.
Who is paying for the buildout?
Some of the early investment is supply-side spending: hyperscale cloud companies, technology providers and other infrastructure operators build capacity ahead of customer demand. Businesses may then pay for that capacity through cloud usage, AI platforms or managed services instead of buying and operating servers themselves. In other cases, an enterprise invests directly in equipment or facilities.
That distinction matters when interpreting market growth. A rising global IT-spending total does not mean every employer is expanding its own data center, and it does not imply that ordinary consumers will see an equivalent change in their personal technology bills. Spending may flow through providers and show up in service fees, subscriptions or enterprise contracts. Nor does a forecast show whether the investments will earn an adequate return.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →What the trend means for budgets and financial decisions
For a household, the Gartner forecast is context about a large technology market—not a reason by itself to change a budget, buy a particular stock or assume an AI-related investment will pay off. Market spending forecasts do not establish a vendor’s revenue, profitability, valuation or future share price. Those depend on company-specific results, competition, costs and expectations as well as market demand.
For business and IT-finance leaders, the same caution applies at the budget level: global spending growth is not a recommended increase for any one organization. A more defensible decision starts with workload demand and total cost, then compares ways to obtain capacity:
- Define the job. Separate training and fine-tuning from batch or real-time inference, data preparation, evaluation and other workloads. Their performance and capacity needs differ.
- Estimate sustained use. An owned cluster can make sense when demand is predictable and utilization is high; intermittent or uncertain demand may favor cloud or managed services. Compare actual contract terms rather than assuming either option is inherently cheaper.
- Count the whole system. Include accelerators, host systems, memory, storage, networking, data transfer, power, cooling, facility costs, software, security, monitoring, backup and staff time. A low GPU-hour rate does not capture total cost.
- Check physical and operational capacity. Confirm power availability, cooling, networking, accelerator supply, delivery lead times and the team’s ability to operate the environment before committing funds.
- Use scenarios and checkpoints. Model demand, utilization, energy costs and possible delays. Tie expansion or longer commitments to measurable business value and cost per useful output, not just a vendor’s capacity offer.
- Preserve options where practical. Cloud and managed services can speed access and reduce operational burden, but consider regional availability, quotas, data-transfer fees, provider-specific tools and minimum commitments. Colocation can offer facility capacity while leaving hardware operations to the customer; owning equipment requires both capital and in-house capability.
Constraints and risks behind the growth
Demand is not the only factor shaping infrastructure spending. Grid connections, transformers, data-center construction, cooling, semiconductor and memory supply, networking capacity, skilled labor and data-residency rules can limit when and where capacity becomes available. Even where a provider advertises AI infrastructure, a customer may encounter regional shortages, quotas, provisioning delays or data-transfer costs.
There is also utilization risk. Expensive hardware can sit idle, or fail to deliver expected throughput because of scheduling gaps, memory limits or slow data pipelines. Inference—the repeated use of a model after training—can become a significant recurring expense as applications scale. Gartner’s October 2025 forecast for AI-optimized IaaS put 2026 spending at $37.5 billion and estimated 55% would support inference; those are forecasts, not audited actuals. This helps explain why buyers should evaluate inference costs separately from training costs. See the Gartner IaaS forecast.
Recommended Free Tools
Gartner’s February 2026 outlook acknowledged concerns about an AI bubble while still projecting rapid infrastructure growth. That is not proof either that a bubble exists or that the buildout will be profitable. Providers may continue investing in expected future demand even if some applications disappoint. The sensible reading is narrower: Gartner expected strong spending, while the pace, availability and financial returns remain uncertain.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
