TSMC projects the global semiconductor market will reach $1.5 trillion by 2030, with artificial intelligence the primary engine of growth. The forecast from the world's largest contract chipmaker puts a hard number on what has, until recently, been described in vaguer terms across the industry. It also implies a roughly trillion-dollar expansion from where the market stood at the start of the decade.
TSMC fabricates the leading-edge silicon behind nearly every major AI accelerator on the market, including parts designed by NVIDIA, AMD, Apple, Broadcom and a growing list of hyperscaler custom chips. When the company puts a number on the decade, it is effectively pricing in its own order book — and the order books of every customer racing to build out AI infrastructure.
The $1.5 trillion figure represents a doubling of the semiconductor industry from its pre-AI baseline in roughly a decade. Industry revenue had historically grown in the mid-single digits annually, with cyclical swings around memory pricing. AI workloads, and the specialized accelerators and high-bandwidth memory they require, have rewritten that growth curve.
Key facts
- 01TSMC projects the global chip market will hit $1.5 trillion by 2030.
- 02AI demand is the primary growth driver behind the forecast.
- 03TSMC is the world's largest contract semiconductor manufacturer and a key supplier to NVIDIA, AMD and Apple.
TSMC has been the most visible beneficiary so far. Its advanced packaging capacity — particularly CoWoS, the technology used to stitch GPU dies to HBM stacks — has been sold out for multiple quarters, with the company repeatedly expanding capacity to keep pace with demand from NVIDIA and others. The 2030 forecast suggests TSMC sees that demand profile holding rather than reverting.
“TSMC expects the global semiconductor market to reach $1.5 trillion by 2030, with AI accelerators as the dominant growth driver across foundry, advanced packaging and HBM.”— Jaeden Schafer
The projection lands as hyperscalers continue to commit unprecedented sums to AI infrastructure. Microsoft, Google, Meta and Amazon are collectively on track to spend hundreds of billions of dollars on data centers and accelerators across 2025 and 2026, with much of that capex flowing through to TSMC's wafer starts and packaging lines.
It also coincides with a global re-shoring push. TSMC is building fabs in Arizona, Japan and Germany, partly in response to customer pressure for geographic diversification and partly under direct prompting from the US and EU governments. Hitting $1.5 trillion globally implies enough demand to fill those new facilities alongside the company's Taiwan base.
The forecast is not without risk. A semiconductor industry doubling assumes AI capex continues at current intensity, that hyperscaler returns on those investments justify further spending, and that no single customer concentration risk — NVIDIA accounts for an outsized share of TSMC's leading-edge revenue — turns into a demand cliff. TSMC's own forecasts have historically been conservative, but a decade is a long horizon for any chip cycle.
There is also the question of who else captures the upside. Samsung Foundry and Intel Foundry are both positioning to claim AI accelerator share, and the memory side of the equation — dominated by SK Hynix, Samsung and Micron on HBM — will absorb a meaningful slice of any $1.5 trillion total. TSMC's number is a market projection, not a market-share claim.
A doubling of the semiconductor industry to $1.5 trillion would put chips on par with the global automotive market in revenue terms, with AI accelerators as the highest-margin segment of that mix. For the AI industry specifically, it sets a clear ceiling on physical-layer supply: every model training run, every inference cluster, every agent deployment ultimately routes through silicon that TSMC and its peers have to build. The bottleneck for the next phase of AI is no longer just algorithms or data — it is fab capacity, packaging throughput and HBM yield, and TSMC has just told the market how big it expects that bottleneck to grow.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
