Chips & Compute

TSMC's September revenue jumps 54.6% as AI chip demand accelerates

TSMC posted September revenue of NT$511.86 billion ($16.03B), up 54.6% year-over-year, as demand for advanced AI chips from customers like Nvidia and Apple keeps surging.

TSMC’s September sales surge year-over-year as AI drives chip demand
TSMC’s September sales surge year-over-year as AI drives chip demandAI-generated
By Marcus Bennett3 min read

Updated

Why it matters

  • TSMC September revenue: NT$511.86 billion ($16.03 billion), up 54.6% year-over-year
  • Q3 revenue came in around NT$1.49 trillion; full earnings report due next week
  • TSMC committed in September to ASML's High NA EUV lithography machines, joining Samsung
  • Shares closed 1.35% lower ahead of the monthly figures release
  • September sales fell 0.6% versus August and did not set a record high, per correction

TSMC posted September revenue of NT$511.86 billion (US$16.03 billion), up 54.6 percent from a year earlier — the sharpest signal yet that demand for advanced AI chips keeps compounding at the world's largest contract chipmaker.

The monthly figure slipped 0.6 percent from August, according to the company's figures. That month-over-month dip does little to dull the annual growth rate, which remains far above what most of the semiconductor industry reports.

Third-quarter revenue came in around NT$1.49 trillion. TSMC is scheduled to report its full third-quarter earnings next week, a release that will give investors margin detail, segment breakdowns and guidance that the monthly sales numbers do not include.

The company's shares closed 1.35 percent lower ahead of the monthly release.

Why TSMC's monthly number matters for the AI supply chain

TSMC sits at a single choke point in the global AI hardware buildout. It counts Nvidia and Apple among its customers, and both depend on its leading-edge process nodes for their most advanced silicon. When TSMC's revenue accelerates this sharply, it functionally confirms that orders for AI accelerators and flagship mobile processors are still flowing at scale — a read on the health of the entire AI compute market.

That matters because TSMC is the foundry of record for the bulk of high-end AI chips shipped today. Its September result acts as a near-real-time indicator in a market where individual chip vendors report only quarterly. A 54.6 percent annual jump, sustained across the quarter, tells policymakers, investors and competitors that capacity investments in advanced logic remain demand-backed rather than speculative.

What changes with High NA EUV adoption?

September also marked a technology commitment. TSMC said it will use ASML's High NA extreme ultraviolet lithography machines, a key technology for producing more advanced chips. The move puts TSMC in the same early-adopter camp as Samsung Electronics, which had already signed on for the tools.

High NA EUV raises the numerical aperture of the lithography system, enabling finer feature printing on wafers — the mechanical foundation for future transistor scaling. Adoption by TSMC, rather than by a second-tier foundry, matters because it effectively standardizes the next lithography generation across the industry's leading edge.

The stakes are straightforward. Whoever controls access to leading-edge capacity controls how fast AI models can scale in hardware. TSMC pairing record-level monthly revenue with a next-generation tool commitment signals it intends to keep that position.

What to watch next week

The third-quarter earnings report next week is the next data point. It will show whether the September surge translated into profit growth, how capacity utilization looks, and what management signals about 2025 demand. One caveat attached to Monday's figures: an earlier version of the report said September sales hit a record high; that claim was corrected. The year-over-year growth stands, the record designation does not.

For now, the pattern holds: AI demand keeps pulling TSMC's revenue upward at rates the company rarely posted before the current chip cycle began.

Original: pr.tsmc.com

Share this article:

More from Marcus Bennett

Marcus Bennett

Show full bio

Senior reporter covering consumer brands and retail at AI In Context.

149 articles

Related articles

  1. Nvidia Added $150 Billion to Its Buyback. One Strategist Says the Stock Is Still Cheap
  2. Micron Posts $54.23 Billion Quarter as AI Memory Demand Explodes
  3. OpenAI in Talks to Raise $30 Billion at $1.4 Trillion Valuation
  4. OpenAI Raises $122 Billion to Scale Frontier AI Worldwide
  5. Goldman Sachs Sees $1.2 Trillion Big Tech AI Spending Wave by 2027

« Previous article