The IT industry has had both memory shortages and gluts many times before, so there’s an understandable reluctance to get carried away when it comes to memory shortages because, for the most part, they sort themselves out. But perhaps things won’t be quite so simple this time around.
The rapid expansion of artificial intelligence (AI) infrastructure and workloads has exerted significant pressure on the memory ecosystem, shifting manufacturing capacity away from consumer electronics toward high-margin memory solutions to support AI. This led to restricted supply of general-purpose memory modules and drove up prices across the board. IDC noted that the timing of the memory shortage “creates a perfect storm for the PC industry, colliding with the Microsoft Windows 10 end-of-life refresh cycle and the AI PC marketing push”.
Since then, there have been several clear indicators that IDC’s warnings were not hyperbole. There were stories of a possible delay to the launch of the Sony PlayStation 6 until 2028 or 2029 because of those shortages, as well as a possible price increase for the Nintendo Switch 2. Multiple reports quoted Pua Khein-Seng, CEO of Taiwanese electronics company Phison, predicting that the memory shortage would lead to several consumer electronics companies going bankrupt or abandoning product lines this year.
There have also been reports of vendors amending their terms and conditions to enable price changes up to the day of shipment. Then, at the end of February, concrete evidence of the effects of shortages and higher prices appeared when HP reported its first-quarter (Q1) results.
CFO Karen Parkhill revealed that memory chip prices had surged 100% sequentially from the vendor’s quarter one to quarter two. To put this into perspective, this represents a doubling of prices from the first quarter ending 31 January 2026 and the second quarter starting 1 February 2026, with memory costs accounting for 35% of a PC’s total cost, compared to 15-18% in Q1. Parkhill predicted memory prices would continue to increase beyond the end of HP’s second quarter.
In December, IDC warned: “Just as the [PC] industry is seeing a need to add more RAM, it has become prohibitively expensive to do so, even if they can get supply. This will result in higher prices, lower margins or a potential downmix in the amount of RAM in new systems at the worst possible time for this to occur.”
Gurvan Meyer, enterprise analyst at Context, agrees: “There will be a cascading effect of the memory shortage in the next months/year. What we are hearing from our contacts is that the shortages are here for [at least] the next two years.”
He notes that there are no massive increases of production lines or manufactory planned until mid 2027, and that the shortages are affecting not only IT products but appliances in general, adding: “Memory has been seen as a commodity by many manufacturers, but it is now a strategic commodity where manufacturers need to plan the sourcing more in advance compared to what they used to do.”
Following IDC, Gartner predicted a 130% rise in combined dynamic random access memory (Dram) and solid-state drive (SSD) prices by the end of this year. That would push PC prices up by 17% compared with last year. Consequently, the market researcher forecast the decline of PC shipments by 10.4% this year and a change in customer buying patterns with an emphasis on prolonging the lifecycle of a PC.
“This is the steepest contraction in device shipments witnessed in over a decade,” says Ranjit Atwal, senior director analyst at Gartner. “Higher prices will narrow the range of devices available, prompting buyers to hold on to devices for longer, fundamentally altering upgrade cycles.”
The sharp increase in memory prices removes vendors’ ability to absorb costs, making low-margin entry-level laptops non-viable, says Atwal, adding: “Ultimately, we expect the sub-$500 entry-level PC segment to disappear by 2028. In addition, rising AI PC prices will delay the projected 50% market penetration of AI PCs until 2028.”
How to address memory shortages
So, can anything be done to minimise the effect of memory shortages on the PC market? How long are they likely to last? And what can channel partners do to help customers navigate the shortages?
“Memory shortages aren’t new, but the driver behind this one is different,” says Darren Hedley, managing director at Insight UK, adding that the technology sector has reached the point where the “boundless potential” of software is met by the “physical constraints of silicon and electricity”. Manufacturing capacity is now being permanently reallocated toward high-margin AI infrastructure. “This shift has resulted in a structural memory crisis that is fundamentally altering the economics of the IT market,” he adds.
Lance Williams, founder of Offerlogic, agrees that IDC and Gartner have the right to be concerned: “Just a cursory look across the market data and we can see the demand for AI continuing to rise as more users log on for the first time and existing users do more with it, more often, creating scaling demand on the datacentre infrastructure that sustains everything.”
Williams states that he’s already noticing it through his personal use of generative AI (GenAI), saying that it becomes less efficient after 3pm GMT when the bulk of the US comes online and response times on his preferred platform slow. “There is little to be done about shortages,” he says. “As we’ve seen across the years, across all industries, when supply outstrips demand, prices increase and it’s those who have the deepest pockets who tend to win out. [Memory supply is likely to] fluctuate across the year, and every year.”
As we’ve seen across the years, across all industries, when supply outstrips demand, prices increase and it’s those who have the deepest pockets who tend to win out
Lance Williams, Offerlogic
Ian Foddering, vice-president of Europe for SHI, describes the concerns about memory prices as justified because the shortages are significantly affecting refresh-and-replace cycles as well as investment in new datacentre capabilities. This situation is leading to increased reliance on break/fix support, extended warranties and device reuse programmes, with standard configurations “becoming increasingly difficult to maintain, especially when the availability of preferred memory tiers, such as 16GB, 32GB and 64GB, begins to dry up”.
Richard Eglon, CMO at Nebula Global Services, deploys an arboreal metaphor to describe the situation: “In nature, the strongest trees rise because the weaker ones fall away, and every forest instinctively knows that resources flow toward growth, not decay. The same is true in any ecosystem, animal, environmental or technological.
“[When AI workloads exploded], the memory industry behaved exactly like a living system under pressure: capacity, investment and manufacturing naturally shifted toward the ‘strongest sapling’, the high-margin, high-demand memory required to power AI infrastructure.”
This rebalancing has starved the more commoditised memory markets, says Eglon, with PCs left to compete with the booming AI sector for the same finite pool of resources. “And they are losing,” he adds. “What we’re seeing now isn’t chaos, it’s ecology. Every action in a complex system has a cost, and prioritising one part of the habitat means another must sacrifice.”
With so much capacity being devoted to AI, price spikes, supply squeezes and roadmap delays are already baked into 2026 for the PC market, he adds.
Andrew Graham, channel development manager at PFU (EMEA), notes that memory shortages are not new and the industry has historically corrected imbalances over time. “However, the structural nature of AI-driven demand suggests that pressure may be more sustained than in previous cycles. This should not be viewed purely as a supply-side issue. Organisational behaviours also influence storage demand and overall memory consumption.
“At a macro level, supply and demand will eventually rebalance. But at a micro level, organisations have more influence than they might assume. Technologies that accelerate digitisation, such as capture solutions, can appear to contribute to storage growth. In reality, when deployed strategically, they can reduce unnecessary expansion”.
While he accepts that pressure could remain for some time if AI-driven demand continues at pace, he adds: “Markets adapt. Investment follows demand and capacity expands. Rather than focusing purely on duration, organisations should consider their exposure to volatility. Those that manage information lifecycle and storage strategically will be better positioned regardless of how long this cycle persists.”
Counter shortages with early action
Hedley at Insight UK, believes the biggest mistake organisations can make in trying to minimise the effects of the shortages is reacting too late. “Buying during peak pricing is always the most expensive option so, instead, organisations should take this moment to rethink refresh strategies,” he says.
Many organisations follow rigid, calendar-based device refresh cycles, but a performance-based approach is far more resilient. “Extending the life of devices that still meet performance standards, planning purchases earlier and staying flexible on specifications or OEM choice can significantly reduce cost pressure during supply constraints,” adds Hedley.
Roger Cummings, CEO at Peak:AIO, suggests that channel partners should focus on two core strategies to try to mitigate the memory shortages that will persist throughout 2026. The first is to try to get ahead of refresh cycles by calculating customers’ hardware refresh requirements 12 to 18 months ahead and locking in allocations “before the market does it for you”. It should be noted that this strategy might not be quite so effective if more vendors choose to amend their terms and conditions to allow price increases to be applied up to the point where the product is shipped.
The second is to help customers to get more out of the hardware they already have. “Constrained supply is the perfect forcing function to have a software conversation,” Cummings says. “Many customers are running inefficient, outdated software on infrastructure that could perform significantly better with the right upgrades, or outright replacement. [This can be achieved by] trimming the bloat that consumes memory and I/O unnecessarily, while introducing leaner, more capable approaches that extract more from existing assets.”
Foddering at SHI believes that channel partners and systems integrators can help customers to navigate the shortages by helping them to adopt more flexible procurement strategies, adding: “Yes, they’ll begin to experience budget volatility as pricing and availability shift between order and delivery, but the channel can advise on how to offset these.”
For example, partners can advise customers on how to create a memory-dependent inventory of hardware to determine where the shortages are liable to bite the hardest. “Deployments can then be prioritised by identifying heavy versus light dependency users and workloads to protect higher-memory configurations where performance is business-critical,” says Foddering.
The inventory should include a shared view of refresh timelines, supply risks and budget triggers, as well as a runbook to decide what to accelerate, what to delay and what to standardise when the situation demands.
At a time when prices are high and availability scarce, buying early offers no guarantees. Therefore, IT leaders must shift from a ‘buy now or wait’ mentality to one of ‘What’s essential, and how do we secure it?’
Ian Foddering, SHI
“Adapting to this new normal is crucial because these shortages will place strain on other areas of the business too,” Foddering adds. “If replacements and spare devices aren’t available, it falls to IT to pick up the slack. They’re likely to see higher break-fix support volumes, and need to maintain and update devices for longer, which means they’ll have less time for digital transformation or strategic work, slowing business growth.
“This isn’t a timing issue. At a time when prices are high and availability scarce, buying early offers no guarantees. Therefore, IT leaders must shift from a ‘buy now or wait’ mentality to one of ‘What’s essential, and how do we secure it?’”
Partners that offer value-add assessments of device and infrastructure estates, as well as real-time performance tracking and digital employee experience monitoring, “will be well-positioned to help teams make the transition, and in doing so make their organisations more resilient to future shortages”, he adds.
Offerlogic’s Williams says that device-as-a-service (DaaS) models, with monthly recurring revenue, could be a useful way for partners to steady their predictable revenues while supporting customers by placing the onus of the device performance and associated costs on the supplier of the DaaS.
In addition, channel partners that claim to have consultative approaches to their end user computing or workspace businesses need to live up to their billing. “This is because, if prices continue to rise, the choice of the ‘right’ device becomes more important and so those channel partners with a solid, consultative workspace business will stand out in their support for their customers navigating their way through the turbulent waters of the memory market,” says Williams.
Another possible alternative could be refurbished PCs. Context recently reported that refurbished PC unit sales through distribution rose 7% year on year in Q4 2025 across Europe’s five largest markets: Italy, the UK, Germany, Spain and France.
“Our latest analysis shows second-life computing moving decisively into the mainstream, with the UK emerging as the fastest-growing market in Europe,” said Jacky Chang, ESG specialist at Context. “As price pressure, constrained supply and sustainability priorities converge, retailers face a narrow window to recalibrate their strategies.
“The trend was towards higher-spec units with higher values. In a market with locked-in supply constraints, these can be an attractive solution for retailers and consumers looking for a solid workaround.”
In the meantime, Gartner has warned that partners and vendors will be forced to choose between eroding margins through lower prices or accepting a reduction in sales when they raise prices – and they don’t have much time to decide.
As Atwal notes: “Device vendors and channels face a critical window in the first half of 2026 to optimise pricing and protect margins before component inflation compresses profitability from the second quarter onwards.”
