A technology enterprise dedicated to Starlight China core engineering, integrating the development/design and industrialization of digital multimedia chips

In 1999, many doctoral entrepreneurs from Silicon Valley founded Zhongxing Microelectronics Co., Ltd. in Beijing Zhongguancun Science and Technology Park, and launched and undertaken the national strategic project - "Starlight China Core Project", focusing on the development, design and industry of digital multimedia chips. change.

In the past twenty years, "Starlight" digital multimedia chip products have been successfully introduced to the global market and are widely used in the rapid growth of multimedia applications such as personal computers, broadband, mobile communications and information and home appliances. Product sales have covered Europe, the United States, and In 16 countries and regions including Japan and South Korea, customers include a large number of domestic and foreign companies. In 2005, it became a Chinese chip design company with independent intellectual property rights listed on Nasdaq.

In response to the development needs of the national intelligent security monitoring network and the Internet of Things industry, Zhongxing Microelectronics has formulated national standards for security prevention video surveillance digital audio and video codec (SVAC), which has contributed to safeguarding national security and production safety, and has made a breakthrough in a China-based industry. A distinctive high-tech development path.

In early 2016, Zhongxingwei launched SVAC video codec SoC integrated with neural network processor (NPU), allowing intelligent analysis results to be encoded at the same time as video data to form a structured video code stream. This technology is widely used in video surveillance cameras, opening a new era of intelligent security surveillance. The independently designed embedded neural network processor (NPU) adopts the "data-driven parallel computing" architecture, which is specially optimized for deep learning algorithms. It has high performance, low power consumption, high integration, small size and other characteristics, which is particularly suitable for The demand for front-end intelligence of the Internet of Things.