AI Is Making Telecom More Expensive, Warn Ericsson and Nokia

Why It Matters

Higher component prices threaten telecom operators’ cost structures and could translate into steeper consumer tariffs, while compressing margins for the industry’s leading equipment makers.

Key Takeaways

• AI demand spikes memory chip profits, Samsung $37.7B operating profit.
• Ericsson and Nokia face ASIC price hikes, threatening RAN margins.
• Telcos may need to pass higher equipment costs to consumers.
• Nokia's optical networking accounts for 18% revenue, margin down to 6.7%.
• Supply lead times extend across chips, prompting price renegotiations.

Pulse Analysis

The AI boom is reshaping semiconductor demand far beyond data‑center servers. Advanced 3‑nm and 5‑nm wafers, once the domain of flagship smartphones, are now essential for the ASICs that power 5G radios and baseband units. As TSMC prioritises high‑margin AI workloads and consumer giants like Apple and Nvidia, telecom equipment makers are pushed to the back of the queue, driving up unit costs and elongating lead times. This mismatch between supply and AI‑driven demand has turned memory chips into a cash cow for manufacturers, with Samsung posting a $37.7 billion operating profit and SK Hynix and Micron enjoying double‑digit growth.

For Ericsson and Nokia, the price surge hits at a vulnerable point. Their networking divisions already operate on thin margins—Nokia’s optical‑networking business slipped to a 6.7% operating margin, and Ericsson’s RAN profitability is under pressure from a weaker dollar and slower 5G rollout. Both firms have begun renegotiating contracts with carriers, seeking cost‑pass‑through mechanisms to protect earnings. The pressure has also accelerated workforce reductions, with Ericsson trimming its staff by over 17,000 since 2022 and Nokia cutting thousands more, underscoring the financial strain.

The ripple effect reaches end‑users. If carriers absorb higher equipment costs, they are likely to reflect them in subscription fees, especially as network upgrades accelerate to support AI‑intensive services. Operators may mitigate exposure by diversifying vendors, extending equipment lifecycles, or leveraging emerging 2‑nm processes that TSMC expects to commercialise soon. Nonetheless, the current chip crunch illustrates how AI’s appetite for silicon can reverberate through the entire telecom ecosystem, reshaping pricing dynamics and strategic planning for years to come.