According to theloadstar.com, Taiwan Semiconductor Manufacturing Company (TSMC) has raised its full-year 2026 revenue forecast to growth of slightly above 40% in US dollar terms, citing sustained, multi-year demand for AI infrastructure — a trend it expects to continue through 2029 or 2030.
Record Q2 performance and upward revision
TSMC reported second-quarter 2026 revenue of NT$1.27 trillion ($40.2 billion), a 36% year-on-year increase. Net profit surged 77% to NT$706.6 billion. The company attributed the results to exceptionally strong demand for high-performance computing chips — primarily AI accelerators and GPUs — which accounted for 66% of Q2 revenue, up from 22% for smartphones.
The firm also increased its planned 2026 capital expenditure range to $60 billion–$64 billion, up from a prior guidance of $52 billion–$56 billion. CEO CC Wei stated that investment over the next three years would be significantly higher than in the previous three-year period.
AI demand reshaping global air cargo flows
Consultancy Aevean reported that data centre–related air trade grew 42% in 2025, driven by a 65% jump in GPU and AI accelerator shipments and a 70% rise in networking equipment. This added approximately 170,000 tonnes to US air imports in Q1 2026 alone — equivalent to about 52 fully loaded widebody freighter flights per day.
TSMC’s own geographic sales data shows 78% of its Q2 revenue came from North America, underscoring the dominance of US-based hyperscalers and chip designers in driving AI infrastructure investment. Yet manufacturing and logistics footprints are increasingly distributed: TSMC is expanding advanced-node capacity in Taiwan, Arizona, and Japan, while Germany will host mature-node production for automotive and industrial applications.
Supply chain implications across tiers
Morrison Express CEO Asok Kumar confirmed widespread capacity constraints across the AI supply chain, telling theloadstar.com earlier in 2026 that many suppliers were “booked out until end of next year, some even until end of ’28… and many are saying this will continue till 2030.”
He noted that the busiest freight corridors reflect this geography: intra-Asia remains the top lane, followed closely by the transpacific route. TSMC is also investing heavily in advanced packaging, generating new cross-regional flows of high-value semiconductor equipment, materials, and components across Asia and the Pacific.
The surge is already affecting adjacent markets. Counterpoint Research reported global smartphone shipments fell 11% in Q2 2026 — their lowest Q2 level in 13 years — partly due to rising memory chip costs as AI data centres absorb increasing volumes of supply.
From consumer electronics to hyperscaler capex
For decades, air cargo’s high-tech vertical was anchored by seasonal consumer product launches. Now, TSMC’s outlook confirms a structural shift: growth is increasingly tied to hyperscaler capital expenditure rather than cyclical device cycles. As CC Wei told analysts:
“Our customers and customers’ customers, mainly the cloud service providers, continue to provide us with a very strong signal and positive outlook. Thus, our conviction in the multi-year AI megatrend remains very high.” — CC Wei, CEO, TSMC
When asked how long supply constraints would persist, he replied:
“I believe from this day on, all the way to probably 2029, 2030, demand is very strong.” — CC Wei, CEO, TSMC
To validate real-world deployment — not just inventory build-up — TSMC said it is independently tracking the construction and location of new data centres worldwide. This verification effort underscores how deeply AI infrastructure investment has become embedded in physical logistics planning, from chip fabrication to final rack installation.
Source: The Loadstar
Compiled from international media by the SCI.AI editorial team.
