The Transformative Rise of AI Chips in the Semiconductor Industry

TAIPEI, Oct. 7, 2024 — In a world racing towards technological advancement, the generative AI market is reshaping the semiconductor industry in ways we’ve never seen before. DIGITIMES Research has unveiled its latest AI Chips Special Report, emphasizing that the intensive training and inference of generative AI models significantly depend on robust computing resources. The spotlight is now on AI chips, especially as businesses explore hardware solutions that meet these rising computational demands.

The Cloud is Leading the Charge

Much of the innovation in generative AI stems from cloud-based platforms. With advancements in wafer foundry processes, cutting-edge packaging, and the increasing demand for AI servers, cloud data centers are at the forefront of this revolution. Stella Weng, an industry analyst, predicts that from 2023 to 2028, high-end GPUs designed for the cloud will see a staggering compound annual growth rate (CAGR) of 44%, while cloud ASIC accelerators are expected to grow even faster at 52%. The upcoming years are crucial for cloud AI accelerators, but as cloud services refine their infrastructures, the surge in AI server demand may gradually shift toward replacement-driven growth.

Wafer Foundries Rise to the Occasion

In response to the soaring demand for cloud and edge AI chips, wafer foundries are not standingstill. They are ramping up their advanced processes and packaging capacities. Eric Chen, another analyst, estimates a 23% CAGR in the expansion of advanced processes below 5 nanometers within the wafer foundry sector from 2023 to 2028. TSMC, a major player in advanced packaging, anticipates a more than 50% expansion in its CoWoS capacity (Chip-on-Wafer-on-Substrate) during the same period. This push is fueled by the need to enhance AI chip performance and shipping capabilities through innovations in photolithography, transistor structures, and advanced packaging techniques.

A New Era for Edge Deployments

The generative AI landscape is also seeing a shift from cloud services to edge deployments. DIGITIMES Research notes that smartphones and PCs are likely to lead this trend. Jim Chien forecasts that global shipments of generative AI smartphone application processors (APs) could soar to 260 million units in 2024, with potential to hit 800 million by 2028. This would represent a CAGR of 65%. The surge in AI smartphone shipments could result in an estimated 240 million AI-enabled smartphones shipped by 2024, with the trend likely pushing penetration rates to over 50% by 2028.

On the PC front, Joyce Chen highlights that vendors are actively testing the AI PC market in 2024. With advancements in processors leading to improved computational performance and energy efficiency, we’re on the brink of a new wave of growth in AI PCs. By 2024, AI PC processor shipments are projected to reach nearly 50 million units, with forecasts suggesting a rise to 150 million units by 2028—representing a 39% CAGR.

A Future Packed with Opportunities

The evolution of AI applications is set to unlock significant business opportunities, expanding existing markets and revolutionizing current technology and capacity demands. The competition within the chip market is heating up, promising dramatic changes across the semiconductor ecosystem.

For those eager to dive deeper into the influence of generative AI on cloud and edge computing, be sure to check out our AI Chip Special Report. It’s packed with insights on technology development throughout the supply chain, insights on the various AI hardware devices emerging, and a detailed look at market shares and growth trajectories for AI chips.

Stay Connected with DIGITIMES Asia

At DIGITIMES Asia, we’re dedicated to providing comprehensive coverage of the global technology landscape, focusing on supply chains, market trends, and application technologies. Our research delivers valuable insights and networking opportunities for businesses seeking to thrive in this dynamic environment.

The AI Buzz Hub team is excited to see where these breakthroughs take us. Want to stay in the loop on all things AI? Subscribe to our newsletter or share this article with your fellow enthusiasts.