Nvidia's graphics accelerator chip is a very successful product, and has become an extremely advantageous depth learning accelerator (Deep Learning Accelerator) in recent years. DLA), applied to the implementation of deep learning artificial intelligence (Artificial Intelligence; AI) software, greatly expand its application category. Nvidia's decision to design open source, a valuable DLA chip, could help nvidia boost market influence, promote sales of other related products, and avoid monopolistic markets for competitors. According to Edgylabs reports, DLA is an extremely valuable asset for NVIDIA, and DLA chips are widely used, including Internet of things; IoT) devices, various robots (Robot) and self-driving (self-driving car), as well as industrial 4.0 (Industry 4.0) applications, are essential components of many new AI systems. However, competitors in the market, and Nvidia to abandon the short-term chip sales benefits, the strategy to choose the Open-source dla chip design, not only may be a brilliant good chess, and perhaps also to protect the third-party depth of learning technology developers in the future development. Since Linux was first made available to the public in the form of open source software, it has been widely adopted around the world, and the development experience of Linux shows that competition emphasizes short-term benefits and cooperation is beneficial to long-term goals. The market potential of AI is enormous, the potential use is all-encompassing, and each use has corresponding specialized hardware, so the type and quantity of future devices will be very large, such as the mainland is actively in response to its Chinese manufacturing (Made in China) measures to improve the degree of automation, Nvidia believes the number of IoT devices may be as high as hundreds of millions of. So Nvidia has to compete in so many markets or meet all the needs, absolutely Lis. With DLA designed to provide open source, Nvidia should be looking for companies with future needs, and may prefer to choose DLA designs that are free to use, rather than paying for competitors ' products. Furthermore, as AI applications become increasingly widespread, it may also help to promote sales of other NVIDIA products. In addition, the more widespread the use of DLA design in open source, Nvidia's share and influence in the industry and market will also rise. Prof. Michigan State University, a professor at the University of Michigan, argues that providing Open-source dla design is a very smart strategy that will not only weaken competitors ' marketing-specific deep learning chip design efforts, but also prevent other vendors from monopolizing the market, It may even make Nvidia more influential in the depth of its study. In addition, companies that use the Open-source DLA design are not substantially modified, so the final design is likely to be compatible with Nvidia's software tools and other hardware.
