TSMC Predicts $1 Trillion Semiconductor Industry by 2030, Driven by AI Chip Demand

TSMC Vice President of Pathfinding and Corporate Research Dr. Min Cao forecasted in a keynote address that revenue from artificial intelligence (AI) chips will push the global semiconductor industry over…

By Ishika Setia

September 5, 2024

2 min read

Comments

Cao spoke at the IMEC ITF Taiwan technology forum in Taipei last week, noting that TSMC has enjoyed a decade-long “technology wave” (personal computers in 1987, smartphones by 1999 and automobiles from this year) after being founded three decades ago. Cao believes that the ongoing transformative wave would be around AI chips, leading to an eruption in demand for high-performance computing (HPC) chips unlike anything seen before.

More About the TSMC Semiconductor Industry

Advanced packaging, advanced logic technology, and advanced memory were the three critical technologies in this AI boom according to Cao. Logic technology would be used to produce processors, memory chips for massive data storage, and packaging solutions that assemble the bare dies in their final package form. Cao’s perspective mirrors an industry-wide move towards AI chips playing a larger role in semiconductor advancements.

The HPC sector has been a particularly strong performer, with TSMC reporting $39.78 billion in revenue from it for the first half of 2024. HPC products represented 52% of TSMC’s revenue in Q2’24 and experienced a sequential increase of 28%. The figure is up from last quarter, when it was 46%, and suggests the increasing importance of HPC technologies in TSMC’s portfolio.

Cao envisions that by 2030, this will add up to a whopping $1 trillion in revenue — and it’s not just chips for AI products but also smartphones, cars, and Internet of Things (IoT) devices. He expects high-performance computing, including AI chips as a type of that category, will be 40% of the total revenue. Despite the fluctuation in the stock price of NVIDIA recently — a record high was hit and then plunged about 20 percent — overall optimism remains for AI products.