This year’s CES convention has showcased another range of exciting innovations: AI on every chipset, smarter endpoint devices and increasingly savvy robots. The headlines are loud, the demos are sleek, and the future seems just around the corner. But for enterprise CIOs, hardware strategy doesn’t proceed at show-floor speed.
CES reflects what’s possible in technical gadgetry; the job of the CIO is to determine what’s practical. Hardware investment decisions are being made in an environment shaped by economic uncertainty, uneven AI adoption and mounting pressure to prove value from every dollar spent. While employees may thrill to get their hands on the latest laptops and accessories, this is rarely 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 tension neatly: “The past one to two years have been the years of CIO ‘hedging.'”
That hesitancy is not necessarily indecision; for many, it’s strategy.
Modernization is driven by agility, not obsolescence
For years, hardware refresh cycles were predictable: new servers, new storage and new desktops every few years. But in 2026, modernization is less about routine updates and more about enterprise agility. Aging hardware isn’t just old but a risk to performance, easily becoming a drag on speed and flexibility.
Nickolaisen observes this firsthand: “We tend to hold onto our hardware for a long time,” he said, noting that infrastructure such as storage arrays can remain in service well beyond vendor reference lifecycles, even up to 15 years. More important than age, he emphasized, is performance in context: “Because of the age of the infrastructure, it is likely the biggest barrier to enterprise agility. When we need to move fast, the older the legacy system is, the more likely it will get in our way.”
Industry data supports this judgment. According to Gartner, worldwide IT spending is forecast to increase by 9.8% in 2026; however, much of this expansion reflects rising component costs and recurring cloud expenditures rather than broad hardware overhauls. CIOs are under pressure to modernize selectively, prioritizing where infrastructure actually limits adaptability and holds back workflows, rather than chasing every CES novelty.
Strategic restraint is often intentional
If modernization is prioritized for specific bottlenecks, that implies de-prioritization elsewhere. In practice, this means CIOs may delay upgrades in parts of the enterprise that are stable, predictable, or not central to innovation. Nickolaisen framed this as a deliberate choice: “To me, the decision is: Where are my biggest agility pain points?”
For many enterprises, that calculus means extending the life of systems that still meet baseline needs, while shifting investment toward compute and storage that directly enable digital transformation or AI workloads. This is not ignorance of innovation, but a recognition that hardware investment must be tightly tied to business impact.
IDC’s hardware market reporting illustrates this: AI-optimized servers — especially GPU-accelerated systems — saw explosive revenue growth last year, reaching a record $112.4 billion in revenue in the third quarter of 2025. However, the bulk of traditional enterprise servers grew at a more modest pace, often buoyed by cloud adoption rather than on-premises refresh cycles. CIOs are effectively buying capacity where it matters and deferring where it doesn’t.
This approach reflects a theme that has emerged in other areas of enterprise technology: focus not on every emerging capability, but on those that unlock measurable value.
“I have to balance my hardware modernization roadmap and priorities against all of the other demands for technology: new initiatives, solving my data issues, new applications and tools, AI,” Nickolaisen said. “It could be that I only have the budget and resources to take on a small set of infrastructure modernization work.”
AI is redrawing the hardware priority map
Many CIOs have been able to avoid too much hardware investment over the past few years, but AI stands apart as a force that reintroduces urgency to hardware strategy. Applications ranging from generative agents to high-throughput analytics place demands on legacy infrastructure that it was never designed to meet. And while cloud provides a flexible on-ramp for many teams, the sheer scale of AI compute means CIOs must evaluate when and where dedicated hardware yields performance or cost advantage versus public cloud.
“For example, I can live with the current state of my network, but to support newer applications I need to modernize my compute and storage,” Nickolaisen said. For him, the uncertainty surrounding AI is part of a broader market challenge that makes every hardware decision feel riskier.
He outlined several concerns CIOs are weighing simultaneously: “What if someone introduces some whizbang technology that accelerates the obsolescence of my current infrastructure? What if someone acquires an element of my infrastructure and overhauls the licensing model and pricing, and makes it less affordable? What if the entire organization is ‘hedging’ and will not invest until there is more market clarity?”
These questions are especially acute when it comes to AI, where workloads may shift among on-premises infrastructure, specialized hardware and cloud platforms faster than traditional planning cycles anticipate.
CES as inspiration, not obligation
CES gestures toward a future where hardware and intelligence are inseparable. For CIOs, the challenge is not to replicate every floor innovation in-house, but to align enterprise reality with strategic opportunity. The most effective hardware strategies will be those that harmonize between where enterprise agility truly needs support — in data centers, in AI workloads and in systems that underpin customer value — and where legacy systems can be sustained without impeding progress.
“Personally, I have decided that I cannot control the market or technology uncertainty, and so I have to be really good at decision-making and researching the options,” Nickolaisen said. “But at some point I need to make a decision, choose a path and move forward — otherwise, I risk falling behind.”
