Surging AI investment continues, along with a mad dash for AI chips and memory chips, leading to supply chain struggles across the industry.

Supply constraints have driven up prices for Intel’s server products in China by more than 10%, a Reuters exclusive revealed, as AMD customers also experience delays. Reuters sources also revealed that CPU shortages have intensified, which could lead to greater challenges for AI companies and manufacturers.

Intel has warned of delivery lead times of up to six months, according to the source.

China accounts for more than 20% of Intel’s overall revenue, but the sources have said its fifth-generation Xeon CPUs are in short supply, leading to the company rationing deliveries.

AMD has also reportedly informed its clients of supply chain constraints, with some delivery lead times being eight to ten weeks. It said in a statement to Reuters it remained confident in its ability to meet customer demand globally based on its “strong supplier agreements and supply chain” – which includes its TSMC partnership.

Memory shortages and undersupplied AI

This news comes shortly after Intel’s earnings call in January, where the company flagged supply constraints. The company suggested that rapid adoption of AI has led to booming demand for traditional compute.

“In the short term, I’m disappointed that we are not able to fully meet the demand in our markets,” Intel chief executive Lip-Bu Tan told analysts at the time, adding that the company was “working aggressively to grow supply to meet strong customer demand”.

CPU shortages are the result of multiple factors, including the global shortage in memory chips, which has been expected to create impact across the industry and undersupply AI.

When memory prices started to rise at the end of last year in China, customers moved to ramp up their CPU purchases to secure a lower price. The current shortage is being driven by surging AI demand for the cutting-edge chips.

“Firms are shifting production capacity towards higher‑margin AI chips, leaving less capacity for traditional chips used in PCs, smartphones and other electronics,” said Makoto Tsuchiya, senior economist at Oxford Economics. “We expect the current supply–demand imbalance to persist over the next few years, as long as the AI investment boom continues.”

He added: “Within the electronics sector, the polarisation between AI and non‑AI players is likely to intensify. Memory chipmakers will benefit significantly from rising prices, while downstream manufacturers will face mounting input costs.”

‘A monumental year for AI’

Surging demand for agentic AI is also contributing to supply strains, given that these systems require larger processing power.

Speaking at the Cisco AI Summit earlier in the week, Cisco chair and CEO Chuck Robbins stated that 2026 would be a monumental year for AI technology – in particular, it will be the year of agentic applications.

“Whether we’re talking to our enterprise customers or governments around the world, we know we have to embrace this,” he explained. “Many of us believe it’s the biggest transition that we’ve ever seen … and I do believe this will be more revolutionary and it’s moving faster than we’ve ever seen.”

Related stories

Nvidia CEO: AI threat to software industry is “illogical”

Cisco CEO: AI is ‘bigger than the internet’, but warns of ‘carnage’ ahead

Nvidia, ByteDance AI chip deal hinges on Trump-era conditions