Gartner Announces Increase of Global Sales of AI Chips to Rise to 33%

Gartner, Inc., forecasts that the global revenue from AI semiconductors is projected to hit $71 billion in 2024, marking a 33% surge from 2023. Gartner also predicts that AI-powered PC shipments will constitute 22% of total PC shipments by 2024. Furthermore, by the close of 2026, it's anticipated that all enterprise PC acquisitions will be AI-enabled. These AI PCs come equipped with a neural processing unit (NPU), enhancing their performance by allowing for extended operation, reduced noise, and better temperature management. They can seamlessly handle AI tasks in the background, opening up new avenues for integrating AI into everyday tasks. Throughout the forecast period, AI semiconductor revenue is expected to maintain double-digit growth, with 2024 projected to witness the highest growth rate. In 2024, AI chip revenue from computer electronics is estimated to reach $33.4 billion, comprising 47% of total AI semiconductor revenue. Automotive electronics are anticipated to contribute $7.1 billion, while consumer electronics are forecasted to yield $1.8 billion in AI chip revenue.

“Today, generative AI (GenAI) is fuelling demand for high-performance AI chips in data centres. In 2024, the value of AI accelerators used in servers, which offload data processing from microprocessors, will total $21 billion, and increase to $33 billion by 2028,” said Alan Priestley, VP Analyst at Gartner.

Fierce Conflict Between Tech Companies

Amidst this growth, there's fierce competition brewing between semiconductor vendors and tech giants. While there's considerable attention on high-performance graphics processing units (GPUs) for handling new AI tasks, major hyperscalers like AWS, Google, Meta, and Microsoft are all investing in developing their AI-optimized chips. Although chip development entails significant expenses, custom-designed chips offer operational efficiencies, cost reductions in delivering AI services, and increased accessibility to AI applications for users. This trend is expected to persist as the market transitions from development to widespread deployment, as noted by Priestley.