If you’re responsible for IT budgets right now, you’ve probably felt it already. Hardware pricing is creeping up. Lead times are extending. And the prevailing idea that technology gets cheaper year on year, is no longer the case.
Looking further ahead into 2026, this isn’t a short-term blip. A combination of structural market changes means higher hardware costs are likely to be the new normal. The good news is that price hikes don’t have to wipe out your entire IT budget. With the right planning and a more strategic approach to procurement, there are still smart ways to stay in control.
Let’s start with why this is happening; and then, more importantly, what you can do about it.
One of the biggest forces at play is AI - specifically, the explosive demand for memory to power large AI models and data centres.
Manufacturers like Samsung, SK Hynix and Micron are diverting production away from standard PC memory and SSDs towards High Bandwidth Memory (HBM), which is essential for AI workloads. That move has consequences for everyone else.
Samsung has already increased prices on certain memory modules by up to 60%, and analysts expect knock-on effects across the PC and laptop market. IDC forecasts average PC prices rising by 8-15% in 2026, driven largely by this memory shortage.
In simple terms: AI infrastructure is consuming a larger share of production capacity, and traditional enterprise hardware is competing for what’s left.
For decades, Moore’s Law (that the number of transistors on a chip doubles roughly every two years) delivered more performance at a lower effective cost. Today, at the most advanced process nodes, that economic benefit is fading. While transistor density still increases, the cost per transistor is no longer falling. And in some cases, it’s rising.
At the cutting edge - 3nm and 2nm chip manufacturing - costs are rising sharply. The R&D investment, energy consumption and complexity required to produce these chips is enormous. TSMC’s reported price increases of 3–10% on advanced nodes, are designed to cover those costs.
The important change isn’t just financial - it’s about expectations. For the first time in decades, the assumption that hardware will naturally get cheaper no longer holds.
Semiconductors don’t just rely on silicon. They depend on a range of critical raw materials; including Gallium, Germanium and Antimony. These are essential for high-performance chips and advanced electronics.
Export controls and geopolitical tensions have tightened access to these materials, particularly following restrictions introduced by China. Although China suspended its ban on exports to the US in November 2025, the disruption has already had an impact, with estimates suggesting global chip production costs have increased by around 8%.
These aren’t issues that resolve quickly. They introduce long-term friction into supply chains, which feeds directly into pricing.
Finally, the minimum standard has moved.
Individually, each of these factors is manageable. Together, they create a very different hardware market to the one we’ve been used to.
This is the point where many conversations drift into doom-mongering. We think that’s unhelpful. Yes, prices are rising, but there are still smart, proactive ways to protect budgets and improve ROI.
If you already know a refresh is coming in the next 6-9 months, pulling it forward into early 2026 can make a real difference.
We’re currently doing all we can to hold hardware margins steady to protect our clients, but industry-wide contract resets are coming. Locking in pricing sooner can reduce exposure to the sharpest increases.
This isn’t about panic buying - it’s about timing.
One of the biggest risks right now is over-buying “just in case”.
Not every user needs an “AI-ready” device with 32GB of RAM. For many roles, 16GB is the genuine sweet spot. Others may benefit more from faster storage or better thermal performance than raw memory.
A simple workload audit often reveals:
Right-fit technology almost always beats blanket standardisation when budgets are under pressure.
Stretching existing hardware fleets can be sensible, especially with proactive maintenance:
But there’s a tipping point. If an ageing device is costing 20% in lost productivity, that “saving” is actually a hidden tax on the business.
The goal isn’t to avoid refreshes at all costs; it’s to make sure extensions deliver net value, not false economy.
One of the least painful ways to offset rising hardware costs is to look elsewhere in the IT budget.
Many organisations overspend 10-30% on unused or underused SaaS and cloud services. Take time to review your licences. See which subscriptions are actively used, which renew automatically, and where consumption has crept up without anyone noticing.
Reclaiming that “ghost spend” and redirecting it into physical infrastructure is often far easier than asking for new budget approval – and it improves overall financial hygiene at the same time.
In 2026, the organisations that plan ahead will be in a much stronger position than those buying at the last minute. Buying exactly what you need - no more, no less - will matter more than ever. And being open with Finance and the Board about what’s coming will save a lot of difficult conversations later.
At TIEVA, we’re not just here to ship boxes. We work with clients to model their 2026–27 IT spend, understand where pressure points are coming from, and make sure there are no surprises when budgets are scrutinised.
If you’d like a conversation about:
we’d be happy to talk. Sometimes the biggest savings come not from buying more cheaply - but from buying more deliberately.
These challenges aren’t unique, and you don’t have to navigate them alone. Get in touch if you’d like to talk through what makes sense for your organisation.