The 2024 DPU & AI Networking Innovation Conference was recently held in Beijing, recognizing projects that have made outstanding accomplishments in data processing unit (DPU) and AI network technology.

Yusur, a company in Beijing Economic-Technological Development Area (Beijing E-Town), won two prestigious awards at the conference: the Brand Industry Leading Award and the Practice Pioneer Award.

The Brand Industry Leading Award recognizes enterprises that demonstrate outstanding brand strategies and market leadership. The Practice Pioneer Award focuses on cases that successfully apply DPU and AI network technologies to solve practical problems, improve efficiency or create new business models.

After six years of development, Yusur has rapidly risen in the field of DPU chip development. Its new generation of chips is iteratively upgraded every two years. The company was the first in the industry to complete the research and mass production of three generations of DPU chips, becoming a major player in the DPU field and being recognized as a national-level specialized and innovative "small giant" enterprise.

In terms of leading industry innovation, Yusur's third-generation DPU chip, the K2-Pro, is the first domestically produced fully functional DPU computing power chip, paving the way for large-scale production of DPUs in data centers. The agile heterogeneous software development platform, HADOS, has also become one of the most fully adaptable, competitive and widely deployed DPU software platforms in China.

Based on powerful DPU software and hardware collaborative technology, Yusur has successively launched a variety of representative DPU products, including ultra-low-latency network DPU cards, cloud native network acceleration DPU cards and RDMA acceleration DPU cards. These products efficiently empower various industries and have provided superior performance and strong controllability DPU chips and solutions for the construction of underlying infrastructure in fields such as finance, cloud native, intelligent computing, general computing and edge computing.