Qualcomm set up a $100 million fund that will invest in startups working on artificial intelligence (AI). The Qualcomm Ventures AI Fund will invest in companies building on-device AI that will be used in autonomous cars, robotics, and machine learning platforms.

Qualcomm Investment Director Albert Wang will lead the new venture fund.

The fund invested in its first venture — Tel Aviv-based AnyVision, an AI development startup working on face, object, and human recognition. Qualcomm contributed an unknown amount to its $28 million Series A funding round, which was led by Bosch in July.

AnyVision’s technology is built to function on a sensor and operate in real-time. It relies on data acquisition and proprietary algorithms to index and analyze everything that is “seen” by the sensor.

Prior to launching the fund, Qualcomm has previously invested in AI startups out of its investment arm. These investments include to SenseTime, a Beijing, China-based facial recognition company, and Cruise, a self-driving car startup which has since been acquired by GM.

Qualcomm has already begun to integrate AI into its chipsets. In April it launched an IoT platform called Vision Intelligence, which included a family of system-on-chips (SoCs) that use an Ai-engine. The SoCs are equipped with machine learning and computer vision software development kits to help cameras in devices handle tasks like tracking, object detection, classification, and facial recognition.

With the fund, Qualcomm joins a number of other chip vendors investing and working on AI. In May, Intel invested $72 million in 12 startups, five of which are working on AI. The others were cloud, IoT, and silicon technologies companies. Chipmaker Micron, in October, launched a $1 million grant for universities and nonprofit organizations to research AI.

A number of companies have also begun work on AI chips. Alibaba Cloud plans to launch its own line of AI chips and quantum computing processors. Huawei also unveiled two chips optimized for AI this year. AWS at its annual conference this week released a custom chip for machine learning. Intel launched a family of AI chips last year. Nvidia is also a big player in the market, with a number of AI chips and processors.