Baidu Invests Into Beijing-Based AI Chip Maker Qingwei

Chinese artificial intelligence (AI) chip maker Beijing Qingwei Intelligent Technology has received around RMB100 million (US$14.73 million) in an angel round of financing from Baidu Ventures in September last year, reported Chinese media outlet 36kr last week.

Other investors in the round include advertising network operator Focus Media Information Technology, XY Capital, Shenzhen-based Guolong Capital and Hangzhou-based machinery company XIZI United Holding Corporation.

Founded in July 2018, Qingwei Intelligent Technology is a Chinese developer of AI chips with an ultra-low power consumption feature.

The start-up’s major project Thinker is an energy efficient reconfigurable processor with hybrid neural networks, fabricated in 65-nm technology, which has received a patent in China.

Its AI chips, including Thinker I, Thinker II and Thinker S, can be applied to consumer electronics such as home appliances, door locks, wearable devices and mobile hardware in the internet of things sector. It also works on the utilization of intelligent voice control and visual recognition for cyber security, such as smart security cameras.

The company was registered under its CTO Ouyang Peng, a Ph.D. graduate of Tsinghua University, with a registered capital of RMB50.30 million (US$7.37 million).

The company has netted two chief scientists who co-lead the Institute of Microelectronics of China’s top academy and research university Tsinghua University, namely Wei Shaojun, the director and Yin Shouyi, the deputy director of the institute.

Since 2006, Mr. Wei has led a team in conducting reconfigurable computing theory and architecture research over software-oriented chips which are designated to adjust in line with software. Having been divided into smaller modules, the chips can be controlled by multiple different pipelines to vary their functions in real time to fit the requirements of the software.

AI is not only about software, it also requires hardware to support different applications. Large enterprises invest heavily in developing their chips that are designed for AI implementation. Google set the stage for the acceleration of neural network computations with their first tensor processing unit.

Related News