In 2018, the current hottest outlet is still artificial intelligence. The wave of artificial intelligence has engulfed all walks of life, and the industry has begun to come into frequent contact with artificial intelligence, whether it is the Internet of Things, smart hardware, smart home, autopilot, or electronics. The heart of artificial intelligence is the AI chip, in this area, NVIDIA is one of the industry leaders, its influence can not be ignored. After the advent of the AI chip era, NVIDIA has also fully deployed AI. According to relevant research reports, NVIDIA will become the leading supplier of artificial intelligence chips, which is creating the standard in the artificial intelligence computing industry.
However, when the arrival of the AI chip era, both giants and start-ups, have come to the same starting line. Therefore, in the field of AI chip, no one wants to take the opportunity to become the new giants, but the real contenders are not only veteran giants such as NVIDIA, AMD and Google, but also more emerging start-ups, especially from China AI chip business.
AI tuyere chase
Imitation of artificial intelligence, not only the contest between enterprises, but also the national level competition. In present China, artificial intelligence has risen to the national strategy. As early as July 2017, the State Council issued the "New Generation of Artificial Intelligence Development Plan", which states that in 2020, the strategic objective of artificial intelligence in China is that the technology and application should be synchronized with the world advanced level, and the artificial intelligence industry will become a new important one Economic growth, the core industries more than 150 billion yuan scale, driven by the scale of related industries over 1 trillion yuan.
January 18, 2018, the national standardization of artificial intelligence group, expert advisory group held its inaugural meeting. At the meeting, the National Standardization Management Committee announced the establishment of the National Artificial Intelligence Standardization Group and the Expert Advisory Group, which will be responsible for overall planning and coordinating the management of China's artificial intelligence standardization.
The meeting also implemented the spirit of the 19 th National Congress of the Party on promoting the in-depth integration of artificial intelligence and real economy. In addition, "Standardization of Artificial Intelligence Aids Industry Development," "Three-year Action Plan for Promoting the Development of a New Generation of Artificial Intelligence Industry (2018-2020) ", Also known as the white paper. We will promote the standardization of artificial intelligence in an all-round way and promote the development of artificial intelligence industry. Just a day later, the Center for Internet Development and Governance of Tsinghua SEM published Baidu's "Report on the Digital Transformation of Chinese Economy Driven by China's Artificial Intelligence - A Study Report on China's Cognition and Application of Artificial Intelligence" (hereinafter referred to as the "Report").
The report argues that artificial intelligence plays an increasingly important role in digital transformation in many fields. With the rapid updating of technologies and algorithms, new scenarios of artificial intelligence continue to emerge and artificial intelligence has become the most important factor in the development of the digital economy The driving force. As can be seen from policy orientation and capital direction, artificial intelligence has risen to a national strategy. The 19th Congress explicitly proposed to promote the in-depth integration of the Internet, big data, artificial intelligence and real economy.
Abroad AI chip market Heroes hegemony
Chip industry Foreign companies have been the monopoly of the core technology, but to the AI chip era, the traditional chip giants have fought in the field of AI chips, NVIDIA is undoubtedly the undisputed market leader in the AI chip. Due to the rise of artificial intelligence, Internet of Things, smart hardware and so on, the market demand for AI chips is rising all the time, which makes the NVIDIA seem breezy.
As we all know, AI chip emphasizes the deep learning ability, and deep learning for computing speed requirements is very scary. NVIDIA GPU chip just to make a large number of parallel computing processors faster than the CPU ten times or even a few times, which has become an artificial intelligence research team and the necessary business "supplies", almost all domestic and foreign AI venture team are using NVIDIA GPU and platform, one can imagine the profits and benefits.
However, AI chip areas want to dominate the NVIDIA or some difficulties, other giants Intel, AMD and Google also have their own advantages eyeing. In the semiconductor market, Intel has been occupying a "leading" position. However, as the rise of artificial intelligence has led to the world's largest PC components and CPU manufacturers are now posed a serious challenge. In this context, Intel chose to join forces to counter NVIDIA.
December 2017, Intel and AMD pioneered a combination of Intel processors and AMD graphics unit notebook chips, for some time caused the industry to shake. The reason why the two rivals will work together to come together, no doubt caused by the AI chip boom. In addition, the second echelon of Google is also trying to trigger a new subversion.
2017 is the culmination of AI chip boom year, and as early as 2016, Google announced that it will independently develop a new processing system called TPU. Designed specifically for machine learning applications, the system reduces the computational precision of the chip and reduces the number of transistors required to perform each computational operation, allowing the chip to operate at a higher number of operations per second for The carefully tuned machine learning model can run faster on the chip. TPU compared with the CPU, GPU, TPU efficiency increased by 15-30 times, 30-80 times the performance increase. There is no doubt that TPU is a perfect AI chip prototype.
China AI chip makers are on the rise
1.Cambrian
The Cambrian is the first AI chip company in the world to succeed in streaming and has mature products. This step is at the forefront of the world. At present, it owns two product lines of AI terminal IP and cloud high performance AI chip. In 2016, it released the world's first dedicated business-depth learning process - Cambrian 1A. Cambrian team comes from the Institute of Computing Technology, Chinese Academy of Sciences, which is China's first specialized in computer science and technology comprehensive study of national academic institutions, Lenovo, Dawn and a number of high-tech enterprises was born from the Institute, At present, it is also an important shareholder in Cambrian science and technology and a long-term cooperative partner in production, study and research.
2. Deep Kam Technology
According to relevant information, Shen Jian Technology was founded in 2016 by four graduates from Tsinghua University and Stanford University. It is an artificial intelligence technology company focused on deep learning processor solutions. It focuses on deep learning processor and compilation R & D original technology, so that all algorithm developers can easily use. Deep Kam Technology announced in October 2017 AI chip plan, and revealed that its own research and development of AI chip "listening Tao", "Sea of View" will be launched in the third quarter of 2018.
3. Horizon robot
Horizon Robot was founded by Yu Kai, former vice president of Baidu Research Institute in 2015, focusing on the processor chip for studying highly sophisticated neural network-based algorithms. On December 20, 2017, Horizon Robotics announced the launch of two embedded artificial intelligence vision chips for smart driving and smart cameras respectively. Among them, the ITW switch for intelligent driving embedded artificial intelligence vision chip named "journey (Journey) 1.0 processor", the chip supports high-performance L2 ADAS system, simultaneously pedestrians, vehicles, lane lines, traffic signs Card, traffic lights and other 8 categories of targets for real-time detection and identification. The camera-facing vision chip, named "Sunrize 1.0 processor," combines depth-learning algorithms for the horizon and can be used in scenarios such as smart security, smart city, and more.
4. bit mainland
According to the relevant information, Bit Continental registered in 2013, is the establishment of the industry's earlier chip startups, now has more than 1,000 employees worldwide, located in Beijing, Taiwan, North America and other regions. The company currently has billions of chip shipments and has built billions of computational data centers around the world. In November 2017, Botech announced a new brand SOPHON, focusing on the field of artificial intelligence. At the same time, it released the world's first tensor acceleration chip, the BM1680, as well as its SC1 / SC1 + and SS1 smart video analytics servers. In addition, the bit mainland will release the second generation of SOPHON AI chip BM1682 in 2018, computing power will be greatly enhanced, follow-up will also be released third and fourth generations.
Players in the AI chip arena is more than that, and more tech companies and entrepreneurs come in. Artificial intelligence is already recognized as a strategic technology leading the way in the future. In the new round of international competition, all the developed countries in the world want to take the lead. Around the plan and policy of AI, the core technology of artificial intelligence, top talents, Standards and other standards for deployment, accelerate the promotion of artificial intelligence technology and industrial development.
The global technology giant is also constantly increasing capital and resources to seize the commanding heights of artificial intelligence development. As for China's AI chip companies, the multiple stimulations of state policies and capital technologies, the future of China's AI chip industry will be incalculable.
