Recently, the 2025 Huawei Partner Conference-Ascend Artificial Intelligence Partner Summit was held in Shenzhen. The Ascend exhibition area set up at the same time focused on the cutting-edge technological achievements in the field of artificial intelligence and large models. TRS was invited to participate in the summit with TRS Training and Inference All-in-One Machine.

On the spot, TRS demonstrated TRS-GPT LLM Integrated Platform and TRS Training and Inferenceing All-in-One Machine based on Ascend AI computing power. TRS-GPT LLM Integrated Platform integrates TRS's existing data middle platform and capability center, and supports AI application creation and operation, optimization and reinforcement learning around specific business scenarios based on basic large models. Through knowledge base, instruction base, scenario library and application management, user-level intelligent applications can be easily customized to realize full-link closed loop of large-model AI applications. It is applied in media, finance, government affairs, public opinion, public security, patent, open source intelligence and other industries in diversified ways such as intelligent assistant, Copilot, AI Agent, etc., providing rich scene-based business functions and service capabilities.

TRS Training and Inferenceing All-in-One Machine is a large-model training and Inferenceing special equipment built by TRS based on Ascend. The hardware level adopts brand-new Ascend hardware, combining high-computing AI hardware, high-performance operator library, multi-card high-speed interconnection and distributed storage to ensure the high performance, high reliability and high safety of system operation. The software level integrates the data, algorithm, model, application and operation capabilities of TRS-GPT LLM, supports industry version customization, provides model privatization application of the whole process, and truly realizes data element support, AI professional empowerment, industry demand landing and user accurate service.