VirtAI Technology

Established in 2019 within Beijing's Zhongguancun High-tech Park, VirtAI Technology provides data center-level AI computing power virtualization and resource pooling software. The company was founded by Dr. Kun Wang, who serves as CEO, alongside Dr. Fei Chen as CTO and Dr. Mao Zou as Senior VP of Engineering. The leadership team brings extensive experience from prior roles at Dell EMC China Research Institute, with deep expertise in computer architecture, distributed systems, and GPU/FPGA virtualization. Since its inception, the firm has secured significant funding through multiple rounds, including a Series B1 in January 2022, attracting investors like Prosperity7 Ventures, Shunwei Capital, and Hillhouse Capital.

VirtAI Technology targets enterprise clients across various sectors, including internet companies, finance, telecommunications, autonomous driving, and scientific research institutions. Its business model centers on licensing its proprietary software to help these organizations optimize the use of their AI computing resources, thereby reducing the total cost of ownership (TCO) and boosting the productivity of algorithm engineers. The company has established partnerships with Dell EMC, Inspur Information, AWS, and iFLYTEK, among others, and serves clients such as Ctrip.com and China Mobile Migu.

The firm's core product is the OrionX AI computing power resource pooling software. This system creates a unified resource pool from an enterprise's various AI computing assets, including GPUs. A key feature is its ability to decouple AI applications from the physical hardware, allowing applications to transparently use remote computing resources on other machines without code modification and with minimal performance loss. OrionX supports fine-grained GPU virtualization, enabling resources to be dynamically allocated when an application starts and released upon exit, which improves management flexibility and can increase hardware utilization by up to 80%. Additionally, the company offers the Gemini AI Platform, which provides comprehensive AI computing power management and an efficient environment for algorithm development and training.

Keywords: AI computing, GPU virtualization, resource pooling, data center optimization, TCO reduction, OrionX software, software defined computing, GPU sharing, AI infrastructure, enterprise AI, computing power management, Beijing, Kun Wang, Gemini AI platform, HPC, AI accelerator, distributed systems, deep learning infrastructure, cloud computing resources, high-performance computing