This 12 months’s CES conference has showcased one other vary of thrilling improvements: AI on each chipset, smarter endpoint gadgets and more and more savvy robots. The headlines are loud, the demos are smooth, and the longer term appears simply across the nook. However for enterprise CIOs, {hardware} technique does not proceed at show-floor pace.
CES displays what’s attainable in technical gadgetry; the job of the CIO is to find out what’s sensible. {Hardware} funding choices are being made in an setting formed by financial uncertainty, uneven AI adoption and mounting strain to show worth from each greenback spent. Whereas staff might thrill to get their fingers on the most recent laptops and equipment, that is hardly ever the wisest use of IT budgets.
Niel Nickolaisen, chairman of the CIO Council at FC Centripetal and director of strategic engagements at JourneyTeam, captured this stress neatly: “The previous one to 2 years have been the years of CIO ‘hedging.'”
That hesitancy shouldn’t be essentially indecision; for a lot of, it is technique.
Modernization is pushed by agility, not obsolescence
For years, {hardware} refresh cycles had been predictable: new servers, new storage and new desktops each few years. However in 2026, modernization is much less about routine updates and extra about enterprise agility. Growing older {hardware} is not simply outdated however a danger to efficiency, simply changing into a drag on pace and adaptability.
Nickolaisen observes this firsthand: “We have a tendency to carry onto our {hardware} for a very long time,” he stated, noting that infrastructure reminiscent of storage arrays can stay in service properly past vendor reference lifecycles, even as much as 15 years. Extra essential than age, he emphasised, is efficiency in context: “Due to the age of the infrastructure, it’s probably the largest barrier to enterprise agility. When we have to transfer quick, the older the legacy system is, the extra probably it is going to get in our approach.”
Business information helps this judgment. In response to Gartner, worldwide IT spending is forecast to extend by 9.8% in 2026; nonetheless, a lot of this enlargement displays rising element prices and recurring cloud expenditures moderately than broad {hardware} overhauls. CIOs are beneath strain to modernize selectively, prioritizing the place infrastructure truly limits adaptability and holds again workflows, moderately than chasing each CES novelty.
Strategic restraint is commonly intentional
If modernization is prioritized for particular bottlenecks, that suggests de-prioritization elsewhere. In observe, this implies CIOs might delay upgrades in elements of the enterprise which can be steady, predictable, or not central to innovation. Nickolaisen framed this as a deliberate selection: “To me, the choice is: The place are my greatest agility ache factors?”
For a lot of enterprises, that calculus means extending the lifetime of methods that also meet baseline wants, whereas shifting funding towards compute and storage that straight allow digital transformation or AI workloads. This isn’t ignorance of innovation, however a recognition that {hardware} funding should be tightly tied to enterprise affect.
IDC’s {hardware} market reporting illustrates this: AI-optimized servers — particularly GPU-accelerated methods — noticed explosive income development final 12 months, reaching a document $112.4 billion in income within the third quarter of 2025. Nevertheless, the majority of conventional enterprise servers grew at a extra modest tempo, usually buoyed by cloud adoption moderately than on-premises refresh cycles. CIOs are successfully shopping for capability the place it issues and deferring the place it does not.
This method displays a theme that has emerged in different areas of enterprise expertise: focus not on each rising functionality, however on those who unlock measurable worth.
“I’ve to steadiness my {hardware} modernization roadmap and priorities towards all the different calls for for expertise: new initiatives, fixing my information points, new purposes and instruments, AI,” Nickolaisen stated. “It may very well be that I solely have the funds and sources to tackle a small set of infrastructure modernization work.”
AI is redrawing the {hardware} precedence map
Many CIOs have been in a position to keep away from an excessive amount of {hardware} funding over the previous few years, however AI stands aside as a pressure that reintroduces urgency to {hardware} technique. Functions starting from generative brokers to high-throughput analytics place calls for on legacy infrastructure that it was by no means designed to satisfy. And whereas cloud gives a versatile on-ramp for a lot of groups, the sheer scale of AI compute means CIOs should consider when and the place devoted {hardware} yields efficiency or price benefit versus public cloud.
“For instance, I can dwell with the present state of my community, however to help newer purposes I must modernize my compute and storage,” Nickolaisen stated. For him, the uncertainty surrounding AI is a part of a broader market problem that makes each {hardware} choice really feel riskier.
He outlined a number of considerations CIOs are weighing concurrently: “What if somebody introduces some whizbang expertise that accelerates the obsolescence of my present infrastructure? What if somebody acquires a component of my infrastructure and overhauls the licensing mannequin and pricing, and makes it much less reasonably priced? What if your entire group is ‘hedging’ and won’t make investments till there may be extra market readability?”
These questions are particularly acute in terms of AI, the place workloads might shift amongst on-premises infrastructure, specialised {hardware} and cloud platforms quicker than conventional planning cycles anticipate.
CES as inspiration, not obligation
CES gestures towards a future the place {hardware} and intelligence are inseparable. For CIOs, the problem is to not replicate each ground innovation in-house, however to align enterprise actuality with strategic alternative. The best {hardware} methods shall be those who harmonize between the place enterprise agility actually wants help — in information facilities, in AI workloads and in methods that underpin buyer worth — and the place legacy methods could be sustained with out impeding progress.
“Personally, I’ve determined that I can not management the market or expertise uncertainty, and so I’ve to be actually good at decision-making and researching the choices,” Nickolaisen stated. “However in some unspecified time in the future I must decide, select a path and transfer ahead — in any other case, I danger falling behind.”
