Shanghai AI lab's new system creates 1 key chip material

A joint team in China led by the Shanghai AI Laboratory has developed an AI-driven system that automates the production of KrF photoresist, a critical material for chip manufacturing, potentially disrupting the existing supply chain dominated by a few foreign companies.

The breakthrough, according to the lab's official announcement, creates a "closed-loop R&D system" using their "Shusheng" scientific large model, moving away from the "black box" capabilities of a few international suppliers. The project was developed with Xiamen University and the Suzhou National Laboratory as part of China's "New Generation Artificial Intelligence" national technology initiative.

The new system combines AI-based decision-making with automated synthesis to achieve stable, high-purity, and efficient creation of the photoresist resin. This addresses a key vulnerability in China's domestic semiconductor supply chain, as KrF (Krypton Fluoride) photoresist is essential for producing a wide range of chips on mature process nodes used in automotive, consumer electronics, and industrial applications.

This development is a significant step toward China's goal of semiconductor self-sufficiency. By potentially removing a foreign dependency, it could bolster the domestic chip ecosystem, impacting valuations of Chinese semiconductor firms and creating long-term competitive pressure on global material giants like Japan's JSR and Shin-Etsu Chemical.

AI Accelerates Material Science

The core of the innovation is the "AI decision + automated synthesis" framework. The 'Shusheng' large model for science was trained to predict the outcomes of chemical reactions, allowing it to design an optimal synthesis route for the complex polymer resin. This AI-designed process is then executed by automated lab equipment, creating a feedback loop where the AI can learn and iterate, standardizing a process that has historically relied on decades of accumulated, often proprietary, human expertise.

While many recent discussions around AI have centered on workforce reductions, with companies like Block and Cloudflare citing AI-driven efficiencies in layoffs, this breakthrough highlights a different application. It showcases AI's potential to accelerate fundamental research and development, solving complex material science challenges. This aligns with comments from IBM's CEO Arvind Krishna, who noted the company is shifting hiring priorities toward AI-related roles even as it automates other functions.

Reshaping the Supply Chain

The global photoresist market has long been an oligopoly controlled by Japanese and American firms. This new AI-driven method in China introduces a new path to market entry that does not require replicating the decades of trial-and-error development of incumbents. If the system can be scaled to industrial production, it could significantly reduce China's reliance on imports for this critical consumable.

This breakthrough is bullish for China's domestic semiconductor sector, particularly for foundries like SMIC and Hua Hong Semiconductor who are major consumers of photoresist. Conversely, it poses a long-term risk to the market share and valuations of established non-Chinese suppliers who have historically controlled this high-margin segment of the supply chain.

This article is for informational purposes only and does not constitute investment advice.