Chinese AI chip startup sets big target

CAMBRICON Technologies Co, a Chinese artificial intelligence (AI) chip startup backed by Alibaba Group Holding Ltd, said last week that it aims to have one billion devices using its semiconductor intellectual properties in three years.

The move is part of China’s broad push to build world-class AI processor companies that can rival Intel Corp, Qualcomm Inc and Nvidia Corp in the future.

“We aim to account for 30% of China’s high-performance AI chip market in three years. The goal is within our reach,” Chen Tianshi, CEO of Cambricon – which is valued at US$1bil – said.

Founded by Chen, a professor at the Chinese Academy of Sciences Institute of Computing Technology, the Beijing-based startup developed Cambricon-1A, the world’s first commercial chip for deep learning applications, last year.

Huawei Technologies Co Ltd’s Kirin 970 chip, which was developed to power its latest flagship smartphone Mate 10, has used Cambricon’s intellectual property. Cambricon’s technologies have also been used in servers developed by state-owned Sugon Information Industry Co Ltd to deliver faster computing and better reasoning capabilities.

“We will focus on both in-device AI and cloud AI. But we won’t make chips for consumer elec- tronic devices ourselves. We will sell our intellectual properties to hardware makers so that they can better integrate AI into their in-house chips,” Chen said.

According to Chen, the company will scramble to build an ecosystem where its AI chip technologies can be used in the fields of smartphones, robotics, drones, autonomous vehicles and other consumer electronics.

In August, Cambricon announced it had raised US$100mil in series A funding, led by SDIC Chuangye Investment Management, a subsidiary of China’s State Development and Investment Corp.

Other prominent investors included e-commerce giant Alibaba, computer manufacturer Lenovo Group Ltd, and the investment arm of the CAS.

Last Monday, the company also launched three new AI chip intellectual properties to enable low-power-consumption image recognition applications, self-driving and other scenarios.

According to Chen, the new chips unveiled are far better than the traditional general purpose processors in boosting image and speech recognition. The chips also boast higher integration density, making them ideal for many devices.

Sun Ninghui, director of the Institute of Computing Technology at CAS, said Cambricon is playing a pioneering role in the global AI semiconductor sector. More efforts are needed to partner with industrial chain partners, so that China is likely to lead the development of AI in the world, he said.

In July, China unveiled a national plan to build a 1 trillion yuan (US$152.5bil) AI core industry by 2030, and said developing homegrown AI processors is an important part of the ambitious goal. — China Daily/ANN