JMT Technology Forms Strategic Partnership with Hangzhou Company to Implement AI Large Models in Lithography Machine Field
On April 7, 2025, Changzhou JMT Technology Co., Ltd. ("JMT Technology") officially signed a strategic cooperation agreement with a Hangzhou-based company, announcing that both parties will jointly advance intelligent upgrades in lithography machine research, production, and maintenance based on their independently developed AI large model technologies. This collaboration follows the technical path of "from small models to large models, and from private databases to public collaboration," combining cutting-edge international Grok3 algorithms with the Deep Seek inference engine to break through key technical bottlenecks in the lithography machine industry and inject new momentum into the high-quality development of China's semiconductor industry.
JMT Technology, as an enterprise long focused on the localization and technological innovation of lithography machine core components, owns multiple independent intellectual property rights and patent technologies; the Hangzhou company has been deeply engaged in the IT service field for many years, successfully implementing government digital platform construction and cross-border enterprise ERP system development projects, with profound technical integration and complex scenario implementation capabilities. This joint effort aims to explore new paradigms of intelligent solutions through deep coupling of AI large model technology with the lithography machine industry.
According to the agreement, both parties will implement technology in phases:
1. Small model verification and private domain data optimization: Relying on the private domain lithography process database accumulated by JMT Technology, combined with the Hangzhou company's experience in algorithm engineering deployment, they will build lightweight models for segmented scenarios to first solve high-frequency problems such as equipment debugging and fault location.
2. Large model integration and public collaboration: Based on Grok3's multimodal data processing capabilities and Deep Seek's efficient inference architecture, they will gradually break through data barriers across the entire chain of R&D, production, and maintenance, creating an AI large model platform covering the entire lifecycle of lithography machines, achieving automation of complex tasks such as process parameter optimization and defect intelligent detection.
3. Industry ecosystem co-construction: By building a public technology middle platform, they will promote model capability sharing among upstream and downstream enterprises in the industry chain, lower the technology application threshold for small and medium-sized manufacturers, and accelerate the overall intelligent transformation of the industry.
As global competition in the semiconductor industry intensifies, lithography machines, as core equipment for chip manufacturing, are crucial for industrial chain security. This cooperation will not only improve equipment yield and maintenance efficiency but also help China's semiconductor enterprises form collaborative innovation synergy by building an open AI technology ecosystem. According to project leader Mr. Li, the first phase model will complete gray-scale testing on a leading customer's production line within six months.