Can Intel Cope with Demand for Next-Gen AI Chips?

Share this article
Share this article
Prioritise Us on Google
Intel's shares have plummeted 13% despite progress in chip manufacturing and AI-driven sales. (Credit: Getty)
Despite making headway in AI chip sales and data centre revenue, investor confidence in Intel appears to be wavering as chip yields remain in flux

Intel has posted better-than-expected earnings for Q4 2025, but its cautious forecast and persistent supply constraints have sent shares down 13%.

Despite making headway in AI chip sales and data centre revenue, investor confidence appears to be wavering as chip yields and foundry expansion remain in flux.

Forecast falls short as chip supply holds back Q1

Intel has outlined revenue expectations between US$11.7bn and US$12.7bn for the current quarter, paired with breakeven earnings. These figures fall below analyst estimates, tempering optimism around the firm’s manufacturing ambitions.

Chief Financial Officer David Zinsner explains that the outlook is weighed down by supply chain constraints, pointing to limitations in chip output.

“We don’t have the supply we need for Q1, but that should improve by the second quarter,” he told CNBC.

David Zinsner, CFO at Intel. Picture: Intel

Supply remains a recurring challenge for the company. Intel continues to grapple with chip production issues, aiming to reclaim ground from key competitors like Taiwan Semiconductor Manufacturing Company (TSMC) and Samsung. Both firms have outpaced Intel in advanced manufacturing technologies in recent years, placing extra pressure on Intel to stabilise its output as demand for next-generation chips rises.

David attributes the subdued start to the year to seasonal factors paired with limited production, a combination that echoes long-standing issues in Intel's supply chain. The mismatch between demand and available supply persists even as the company accelerates plans to improve manufacturing timelines.

Leadership shift raises expectations

Lip-Bu Tan stepped in as Intel’s new CEO March 2025, bringing with him a strategy focused on regaining market share, reinvigorating the company’s foundry business and scaling competitiveness across the board. This earnings report serves as a major milestone early in his tenure.

Lip-Bu has acknowledged progress in manufacturing, specifically in yield, but admits performance is still not at target.

“Our yields are in line with our internal plans,” he says, “but they’re still below what I want them to be.”

Lip-Bu Tan, CEO at Intel. Picture: Intel

Despite leadership change, investor sentiment heading into the report was high. Intel’s stock doubled over the past 12 months, buoyed by enthusiasm around its 18A manufacturing node – a next-generation chip production technology – and hope that Intel could secure large customers for its growing foundry services.

That optimism has cooled, with Lip-Bu having stopped short of confirming new partnerships – instead stating that Intel is “working aggressively” to boost output in order to match market demand.

Without firm commitments from major chip customers, questions arise around whether Intel’s foundry operation can become a sustainable revenue engine. The 18A node, billed as Intel’s path back to manufacturing leadership, still needs proof points beyond technical promises.

AI and data centres lift performance but PC chips drag

Intel’s Data Center and AI group posted 9% year-on-year growth, reaching US$4.7bn in quarterly revenue. This increase follows stronger demand for server chips designed for AI workloads, an area where Intel’s Xeon processors play a central role.

Lip-Bu highlights the strategic value of these chips in AI infrastructure, stating: “Our central processing units are becoming more critical to systems built for AI.”

In this market, performance per watt matters. Intel’s ability to combine central processing unit (CPU) performance with built-in AI acceleration aligns with cloud providers’ and enterprise clients’ growing focus on energy-efficient computing. This integration could provide Intel a foothold in a space where graphics processing units (GPUs) – such as those produced by NVIDIA – typically dominate.

Youtube Placeholder

However, the strength in AI chips does not offset weakness elsewhere. Intel’s Client Computing Group, which includes chips for laptops and desktops, reported a 7% fall in revenue to $8.2bn. This segment reflects continued softness in consumer demand for personal computers, with pandemic-era highs giving way to sustained contraction in unit shipments.

The imbalance across Intel’s business segments shows that while AI brings upside, it is not yet large enough to counter the drag from other parts of the portfolio.

Outlook hinges on yield and foundry execution

Intel’s trajectory now depends on whether Lip-Bu and his team can raise chip yields and secure high-volume customers for its foundry business. Both goals are key if Intel is to move from promise to performance.

Supply constraints and cautious forecasts suggest that recovery remains an ongoing process, rather than an immediate turnaround. While the AI and data centre momentum points to a clear strategic direction, execution in manufacturing remains the determining factor in Intel’s next phase.

Until it delivers stronger consistency in output and lands external clients for its 18A node, investor caution is likely to persist. For now, Intel’s story remains one of potential – still waiting to match promise with delivery.

Company portals

Executives