TSMC raises prices across all advanced nodes: impact on Nvidia, AMD, Apple

What happened?

TSMC has notified its customers that it will apply price increases across its entire portfolio of advanced nodes, according to a report from specialized media Culpium on June 23. The increases, which vary by customer, node, and product, range from 5% to 10% and include nodes from 7nm up to the most advanced 3nm and 2nm. The company confirmed the strategy without giving specific figures, stating that it 'does not comment on pricing' and that its 'pricing strategy is not opportunistic but strategic.'

According to data from the first quarter of 2026, advanced nodes (7nm and above) accounted for 74% of TSMC's wafer revenue, with the 3nm node contributing 25%. This means the increase affects nearly three-quarters of the Taiwanese manufacturer's business. The Culpium report, reported by Tom's Hardware, notes that the increases have already begun to be applied in some cases, while other customers have been informed to incorporate the higher cost into future purchase orders. The move contrasts with previous statements by CEO C.C. Wei, who had promised to keep prices stable during the April 2026 earnings conference, stating that 'it would be a long time before we could satisfy customer demand' and that he would refrain from implementing increases. However, inflationary pressures, the costs of building new plants in the US, Japan, and Europe, and the imbalance between supply and demand for AI chips have led TSMC to harden its stance.

Why is it important?

TSMC is the world's most advanced chip manufacturer, producing processors for Apple (A17, M4), Nvidia (Blackwell GPU), AMD (Ryzen, EPYC), and Qualcomm (Snapdragon). Any increase in wafer costs directly translates to final chip prices, and eventually to electronic devices and AI servers. The decision to raise prices on mature nodes like 7nm, which is still widely used in automotive, networking, and industrial applications, indicates that TSMC seeks to monetize its production capacity in a context of high demand and rising costs due to the global expansion of its factories.

Historically, TSMC has avoided broad price increases to maintain relationships with its key customers. The last significant increase occurred in 2021-2022, when the global chip shortage allowed increases of up to 20% on some nodes. However, at that time the company focused on mature nodes; now, by including the most advanced nodes (3nm and 2nm), the signal is more aggressive. According to Tom's Hardware, previous reports from Taiwanese media pointed only to increases on 3nm, but Culpium revealed that the measure covers all advanced nodes, including 5nm and 7nm. This is relevant because 7nm remains a high-volume node for automotive and industrial applications, sectors already facing cost pressures.

The impact on the semiconductor market is twofold: on one hand, it reinforces TSMC's dominant position as the undisputed leader in advanced manufacturing; on the other, it threatens to increase inflation across the entire electronic supply chain, from AI servers to smartphones. Analysts estimate that TSMC's gross margins, which hover around 53-55%, could improve by 1 to 2 percentage points with these increases, justifying investor pressure for the company to capitalize on its technological advantage.

Consequences for the industry

TSMC's customers face tighter margins. Nvidia, which relies almost exclusively on TSMC for its AI GPUs, could see its gross margins reduced, although its market dominance allows it to pass the cost on to hyperscalers like Microsoft, Amazon, and Google. Apple, with its enormous bargaining power, could absorb part of the increase, but analysts expect iPhone and Mac prices to rise slightly. AMD and Qualcomm, with less pricing power, could suffer more. Broadcom and MediaTek, also mentioned in the Culpium report, face similar challenges.

In the long term, this increase could accelerate plans by some customers to diversify their supply chain toward Intel Foundry or Samsung Foundry, although both still do not match TSMC's capacity and performance on advanced nodes. Intel has announced its Intel 18A process for 2025, but has yet to demonstrate comparable volumes. Samsung, for its part, has had yield issues on 3nm. Therefore, dependence on TSMC will remain at least until 2027-2028. The increase could also incentivize chip designers to optimize their products for older nodes or seek advanced packaging as an alternative, such as using chiplets to reduce reliance on cutting-edge nodes.

In the past, similar TSMC increases led customers like Apple to negotiate long-term capacity agreements, securing preferential pricing. We are likely to see more of these agreements, further consolidating TSMC's position as a strategic partner. However, smaller companies without bargaining power could be forced to raise prices or reduce R&D investment, affecting innovation in sectors like IoT or automotive.

What should readers know?

Consumers will feel the impact with a delay, as supply contracts typically have 6- to 12-month lead times. Prices for high-end smartphones, gaming PCs, and AI servers are likely to increase in the second half of 2026. For example, an iPhone 18 or an Nvidia RTX 5090 GPU could be 3% to 5% more expensive. For investors, TSMC reinforces its pricing power, which could boost its revenue and margins, but also creates uncertainty about future demand if customers choose to reduce orders. TSMC's shares have risen 8% since the announcement, according to market data.

In summary, TSMC's price increase is a strategic move that reflects its dominant position and the structural challenges of the semiconductor industry. The decision will have repercussions across the entire value chain, from chip manufacturers to end users. The key question is whether customers will passively accept the increases or seek long-term alternatives, which could redefine the balance of power in the industry.