en.Wedoany.com Reported - SUPCON Technology Co., Ltd. and Inspur Electronic Information Industry Co., Ltd. recently formally signed a strategic cooperation agreement. The two parties will focus on Autonomous Operation Plants (AOP), promoting deep collaboration between the time-series large model TPT and a diverse computing power base, jointly creating highly reliable, multi-service concurrent industrial AI solutions. SUPCON disclosed this information through official channels on May 18.
The core of this cooperation is to systematically connect SUPCON's process data and mechanism models accumulated in the process industries with Inspur Information's full-stack capabilities in computing infrastructure. TPT is a self-developed Time-series Pre-trained Transformer model by SUPCON, designed for continuous production processes in the process industries. It assists operational decision-making and autonomous control by predicting the changing trends of key process variables such as temperature, pressure, flow, level, and composition. The model has completed deployment verification in multiple major projects—it was first put into operation on the 450,000 tons/year High-Density Polyethylene (HDPE) unit of Zhejiang Petroleum & Chemical Co., Ltd. in September 2025, successfully launched on the methanation synthesis unit of Inner Mongolia Datang International Keshiketeng Coal-to-SNG Company in December 2025, and achieved hot commissioning operation on the 50,000 tons/year caprolactam unit of Hubei Sanning Chemical Co., Ltd. in March 2026. The three deployments cover three typical process industry routes: refining, coal chemical, and fine chemical, verifying TPT's cross-unit migration capability in different process scenarios.
Inspur Information will provide a diverse computing power base for the large-scale training, inference, and engineering deployment of TPT. Inspur Information is one of the world's major suppliers of AI servers and intelligent computing infrastructure. Its AI servers, edge intelligent gateways, and intelligent computing center products serve large AI workloads across multiple industries, possessing technical expertise in liquid cooling, high-speed interconnection, and system energy efficiency optimization. At the signing ceremony, Shao Bo, General Manager of Inspur Information's Northeast Region, stated that Inspur Information is committed to driving industrial upgrades through intelligent computing power, having deployed several important computing power projects in regions like Liaoning, and hopes to continue making efforts in areas such as regional co-construction and artificial intelligence computing infrastructure.
The difficulty of implementing industrial AI has long been an industry consensus. Compared to general-purpose large language models, AI applications in the process industries need to simultaneously meet multiple constraints such as real-time performance, reliability, interpretability, and anti-interference capability. No single-point technology can independently complete the closed loop from data acquisition and model inference to control command issuance. SUPCON possesses a full-stack product system covering underlying instruments, control systems, and industrial software, while Inspur Information provides a computing power support platform ranging from the chip level to the data center level. The technology stacks of the two enterprises naturally couple along the "data—computing power—model—control" chain of industrial AI. Through this cooperation, TPT's model capabilities can undergo large-scale engineering verification leveraging Inspur Information's intelligent computing infrastructure, providing highly reliable, low-latency industrial AI solutions for process industry customers in petrochemicals, chemicals, power, metallurgy, and other sectors.
The vision of an Autonomous Operation Plant is to equip production lines with the capabilities for autonomous perception, autonomous decision-making, autonomous execution, and autonomous optimization, rather than simply replacing manual operations. AOP requires real-time closed-loop integration of process mechanism models, AI inference engines, and underlying control systems, placing extremely high demands on computing latency, data throughput, and system stability. The partnership between SUPCON and Inspur Information aims precisely to open up the vertical link from upper-layer AI models to the underlying computing power infrastructure, providing a technical foundation for AOP to move from single-point pilots to large-scale replication.
SUPCON is headquartered in Hangzhou, Zhejiang Province, and is a leading enterprise in the field of process industry automation and intelligence, with products and services covering process industries such as chemicals, petrochemicals, power, metallurgy, building materials, and papermaking. Inspur Information is headquartered in Jinan, Shandong Province, and is one of the world's major suppliers of AI servers and intelligent computing infrastructure. The cities where both parties are located are core nodes in China's computing power and industrial intelligence layout. Against the backdrop of the global process industry accelerating its embrace of AI and the continuously rising demand for collaborative innovation between models and computing power, the strategic alliance between these two leading enterprises, headquartered in China's core industrial cities, will provide integrated software and hardware technical support for the large-scale implementation of Autonomous Operation Plants.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
