chip industry, there are few entrepreneurial booms, but the craze that AI has once again attracted many technology giants around the world to enter the chip industry, but also allowed us to witness the rise of AI chips. As an important gathering place for AI chip companies in the world, you should be curious who has joined the AI chip tide? Do these people have some common labels?
A little bit of sorting out it can be found that some people have rewritten the history of China's chiplessness and created glory internationally; some have become famous at a young age and are hailed as geniuses, and have been promoted by Enshi and their fellow students to start their entrepreneurial journey; some have been working in the chip industry for many years, exploring new possibilities with their many friends in the AI era; and some have led the company to impact the extremely hot capital market, shining brightly. In the end, their story turned into a typical symbol of China's chip history for more than 30 years.
If you count the schools that Chinese AI chip leaders graduated from undergraduate studies, you will find that Tsinghua University and the University of Science and Technology of China (hereinafter referred to as the University of Science and Technology of China) have supported the super team of China's AI chips. This article will use their undergraduate school as the main line of connecting the full text to understand the entrepreneurial experience and achievements of China's leading AI chips.
University of Science and Technology opened the chapter on AI chips in China
In June 2016, China's first embedded neural network processor "Starlight One" was released. This chip has been prepared for more than 5 years and has been developed for more than 3 years before it was released. The release of Starlight One is Zhongxing Microelectronics, a veteran domestic chip company. This is a company founded by Academician Deng Zhonghan in 1999.
Interestingly, Academician Deng Zhonghan majored in the undergraduate degree at the University of Science and Technology of China. After graduating from the University of Science and Technology of China and entering the University of California, Berkeley, Deng Zhonghan began to have a relationship with chips, obtained a doctorate in electronic engineering, and also obtained a master's degree in economics and management and a master's degree in physics. He is the first scholar to span the three disciplines of science, engineering and business in the 130 years since the school was founded. After graduating from
, Deng Zhonghan chose the chip industry and worked for IBM and SunMicrosystem successively. Later, he founded the semiconductor company Pixim in Silicon Valley.
Deng Zhonghan, academician of the Chinese Academy of Engineering and commander-in-chief of "Starlight China Core Engineering". Picture source: Zhongxing Microelectronics
In 1999, Deng Zhonghan returned to China and registered and established Zhongxing Microelectronics Co., Ltd. in Zhongguancun in Beijing. The "Starlight" series of chips he led to end the history of China's corelessness. In addition, starlight chips were widely used by well-known domestic and foreign companies such as Apple, Samsung, Philips, and Lenovo . At that time, it successfully occupied more than 60% of the computer image input chip market.
However, leading the Zhongxing Micro team to successfully mass-produce NPUs with disruptive data-driven parallel computing architecture is Zhang Yundong, co-founder of Zhongxing Micro Electronics and CTO of Zhongxing Micro Group. He obtained a bachelor's and master's degree from Fudan University , majoring in microelectronics. After graduation, he worked in Realtek, Taiwan for 5 years, and then started a business in Silicon Valley, USA. Later, he co-founded Zhongxingwei with Deng Zhonghan.
Compared with Zhongxingwei, the AI unicorn company Cambrian Technology, the Department of Chinese Academy of Sciences (hereinafter referred to as "Chinese Academy of Sciences", the supporting schools of the institute are University of Chinese Academy of Sciences and University of Science and Technology of China), has a higher attention. Cambrian Technology was founded by two brothers from the Junior Class of University of Science and Technology. Chen Yunji, who was admitted to the Junior Class of University of Science and Technology at the age of 14, thought computers were very interesting and wanted to go to the Institute of Computer Science and Technology of the Chinese Academy of Sciences to study for graduate school in his senior year. But his undergraduate grades are not outstanding. Fortunately, Hu Weiwu, the leader of the Loongson No. 1 development team, saw that he played well in games and had the potential to do scientific research, and he was reluctant to recruit him as a graduate student.
The younger brother Chen Tianshi, two years younger than Chen Yunji, almost followed his brother's footsteps from the junior class of the University of Science and Technology to the Institute of Computing of the Chinese Academy of Sciences step by step. In 2008, the two brothers decided to work together to conduct cross-study research on AI and chip design. But AI was not at the time, and their series of processor research and development technical papers based on AI methods were rejected by top architecture conferences many times.
Cambrian Technology Chen Tianshi (left) Chen Yunji (right)
Under the influence of his mentor Hu Weiwu, Chen Yunji continued to stick to this direction. In 2015, a 20-person R&D team developed a prototype chip for deep learning special processors.
In early 2016, Chen Yunji met Le Jinxin, a partner of Yuanhe Yuandian. After chatting for a long time, Le Jinxin believed that their team would become a famous company in the AI field in just 1-2 years of starting a business. In March, Cambrian Technology was established and Yuanhe Origin became an angel round investor.
In 2017, the world's first SoC Kirin 970, which is integrated with NPU, released by Huawei , is equipped with Cambrian Technology's 1A IP, which has attracted great attention. In 2017, Cambrian Technology announced that it had completed a US$100 million round of financing. After completing a US$100 million round of financing in June 2018, Cambrian Technology's valuation reached US$2.5 billion, quickly becoming a global AI unicorn.
With the support of capital, Cambrian Technology can naturally invest better in product research and development. In May 2018, Cambrian released its first cloud-based AI chip and board, taking the lead in completing the integrated layout of end-cloud and maintaining continuous updates of chips.
February 28 this year. The website of the Beijing Securities Regulatory Commission disclosed that CITIC Securities and Cambrian Technology signed a counseling agreement on December 5, 2019, and planned to be listed on the Science and Technology Innovation Board.
, like Cambrian Technology, may become the first AI chip stock on the Science and Technology Innovation Board, Yuntian Lifei, co-founder Tian Dihong, is also a graduate of the junior class of the University of Science and Technology of China. Tian Dihong, who was admitted to the Youth Class of the University of Science and Technology (Teaching Reform Pilot Class) for the Chemistry Olympics in 1995, graduated from the Department of Electronic Engineering of the University of Science and Technology in 2002. He later chose to continue his studies in the United States and obtained a doctorate in electronics and computer engineering from Georgia Institute of Technology in 2006.
Yuntian Lifei co-founder Chen Ning (left) Tian Dihong (right) Photo source: Beijing News
Yuntian Lifei's other co-founder and CEO Chen Ning is also a Ph.D. in Georgia Institute of Technology in the United States, majoring in electronic engineering. After graduation, he first joined Freescale at the time, and in 2009 he joined ZTE Communications as the technical director of chip design.
During his tenure at ZTE, Chen Ning led his team to achieve many achievements in the field of communications. But he also keenly discovered the blue ocean of visual intelligence and machine learning processor. So, Chen Ning left ZTE, where he had worked for five years, and together with alumni Tian Dihong, founded Yuntian Lifei in August 2014.
In 2015, Yuntian Lifei and Huawei and Longgang District Public Security jointly created the cloud-based Yuntian "Shenmu" system. In two years, it has helped solve more than 2,300 cases and retrieved many missing children and missing elderly people.
It was also in 2017 that Wang Xiaoyu and his family moved from the United States to Shenzhen and joined Yuntianlifei as chief scientist. Wang Xiaoyu studied in the Department of Electronic Engineering and Information Science of the University of Science and Technology of China. From 2005 to 2008, Wang Xiaoyu, who was studying for a doctorate at the University of Science and Technology of China, began to start a business in the field of security. But in 2008, he decided to drop out of school and go to the United States. After obtaining a doctorate in electronics and computer engineering at the University of Missouri Columbia Campus, Wang Xiaoyu became an IT elite in Silicon Valley.
In the early days of Yuntian Lifei's business, Wang Xiaoyu had continuous communication with the founding team, but in 2015 he jumped from NEC to Snap to start his second business. Two years later, he joined Yuntian Lifei without hesitation.
Yuntian Lifei's three overseas returnee doctoral co-founders Chen Ning, Tian Dihong and Wang Xiaoyu focused on the fields of chip, vision and machine learning respectively. Yuntian Lifei's landing scenario has also expanded from AI+security to AI+new policing, AI+new governance, AI+new retail and other fields.
Yuntian Lifei, who has been making rapid progress, completed a round of financing of hundreds of millions of yuan in March 2019. Yuntian Lifei, which has gone through four rounds of financing, is considered by the outside world to prepare for listing on the Science and Technology Innovation Board. In November 2019, Yuntian Lifei officially released the first AI chip DeepEye1000, which has been struggling for three years. The person who led Yuntian Lifei to develop this product was Li Aijun, who joined Yuntian Lifei in February 2017, serving as vice president and head of chip product line. He graduated with a master's degree from Xi'an University of Electronic Science and Technology of Electronic Science and Technology of China.
Zhaoguan Electronics is not as famous as Cambrian Technology and Yuntian Lifei, but Zhaoguan Electronics is also a typical startup company in the University of Science and Technology. In 1997, Feng Xinpeng, who was very interested in circuits in high school, chose the Department of Electronics of the University of Science and Technology of China. Immediately afterwards, he went to the University of Southampton to study for a master's degree in microelectronics. Two years later, Feng Xinpeng, who graduated from a master's degree, joined SMIC, which was founded shortly after he was engaged in IP design.
Two more years later, Feng Xinpeng quit his job at SMIC and joined AMD. He was promoted to AMD R&D director all the way, leading a team of 100 people from four countries around the world to design 55 CPUs and GPU chips.
About 2012, Feng Xinpeng and Zhou Ji, who was then the chief visual architect of AMD, began to pay attention to the direction of computer vision . Zhou Ji obtained a doctorate from Zhejiang University in 2003, focusing on three-dimensional reconstruction based on images. By 2016, they felt that many opportunities had emerged and they could not continue to wait. Therefore, on March 1, 2016, Feng Xinpeng officially submitted a business plan to the outside world. After 21 days of waiting, he finally received a letter of investment intention.
Zhaoguan Electronics 3 founders took a photo at the 2017 CES exhibition. Zhou Ji (left), Feng Xinpeng (middle), Cai Haijiao (right),
In May 2016, Feng Xinpeng brought Zhou Ji, who had worked at AMD for 13 years, and Cai Haijiao, an alumnus of the University of Science and Technology of China, jointly founded Zhaoguan Electronics. Cai Haijiao has been studying image processing technology when he was in undergraduate degree at the University of Science and Technology of China. When he was a graduate student at the Chinese Academy of Sciences, he was engaged in technical research and product development of infrared cameras. His first job after graduation was to develop products at the well-known semiconductor company Marvell.
However, the three partners have been arguing about the entry point of entrepreneurship for a long time, and the focus of the debate is whether to make products first or chips first. The consensus they finally reached was: first make products, and at the same time accumulate core technologies for manufacturing chips, and then choose the right time before starting the research and development and manufacturing of chips.
They first made glasses that can help blind people "see". After gaining market recognition, in 2018, Zhaoguan Electronics released an AI vision chip N171 that claims to surpass Intel Movidius MyriadX and Nvidia Tegra X2.
In addition to these companies that have been working on AI chips since their inception, they joined the AI chip industry after a few years of starting their business, and are also a representative of AI companies in the University of Science and Technology. Zhou Xi, Wen Hao and Li Jiwei, the three founders of Yuncong Technology, were admitted to the University of Science and Technology of China in 1999 and became roommates.
Yuncong Technology's three founders, Zhou Xi, the left one, Wen Hao, the right two, Li Jiwei, the right two, Li Jiwei,
In 2011, Zhou Xi was introduced to China as an expert in the "Hundred Talents Program" of the Chinese Academy of Sciences. In conjunction with the National University of Singapore, an intelligent multimedia center was established at the Chongqing Institute of the Chinese Academy of Sciences. The newly established multimedia center was seriously lacking in professional talents. So he asked Li Jiwei and Wen Hao for help. Under the leadership of the three, the multimedia center has developed rapidly and has successively developed a series of artificial intelligence systems such as the first domestic face-scanning payment prototype system, intelligent sending, intelligent clothes changing, intelligent cameras, vehicle attribute recognition, and face attribute analysis.
In 2014, Zhou Xi found that the dozens of R&D personnel of the multimedia center could no longer meet the research needs, and there was also a huge gap in operation management and business. In April 2015, Yuncong Technology was officially established and entered a track of rapid development. In September 2018, Yuncong Technology's "High Accuracy Face Recognition Industrial Application Based on Self-developed SoC Chips" was selected as the list of innovative projects for deep integration of artificial intelligence and the real economy announced by the Ministry of Industry and Information Technology.
Interestingly, Zhou Xi and Yun Zhisheng CEO Huang Wei were in the same laboratory while studying at the University of Science and Technology. After graduation, the two fellow students exchanged their entrepreneurial directions. Zhou Xi, who studies voice at the University of Science and Technology of China, is in computer vision, while Huang Wei, who studies images, squeezed into the voice track with his predecessor iFlytek.
Yunzhisheng CEO Huang Wei
However, Huang Wei started his business earlier. He founded Yunzhisheng in 2012 and released the Internet of Things AI chip Swift in May 2018. Not long after, in January 2019, its multimodal AI chip strategy and plan were released.
If the graduates of the University of Science and Technology of China have created the chapter of China's AI chips, the other major faction of China's AI chips, Tsinghua graduates have injected more vitality into China's AI chips.
Tsinghua system injects more vitality into China's AI chips
To talk about Tsinghua system AI chip companies, the most talked about is Shenjian Technology. The company acquired by FPGA giant Xilinx in 2018 was co-founded by "Tsinghua Master and Apprentice Four". Half of the people in the team before the acquisition were from Tsinghua.
Shenjian Technology CEO Yao Song, the supervisor of undergraduate is Wang Yu, associate professor of the Department of Electronic Engineering and deputy secretary of the Party Committee of Tsinghua University. He is a graduate of the 1998th grade of Tsinghua Electronics Department and has rich experience in hardware acceleration research.
In 2015, 23-year-old Yao Song was about to graduate from undergraduate studies. He gave up the offer he had dreamed of studying at Carnegie Mellon University, and started his business with the support of his mentor Wang Yu. In March 2016, Shenjian Technology was established in Beijing.
Shenjian Technology founding team, from left: co-founder Han Song, co-founder and CEO Yao Song, co-founder Wang Yu, CTO Shan Yi, and Yao Song also started a business with Yao Song. Han Song was admitted to the Tsinghua Electronics Department in 2008, and Dr. Stanford was studying at the time. His “Deep Compression” paper won the Best Paper Award at FPGA 2017, the top conference in the field of FPGA chips. Before being acquired, Han Song served as the chief scientist of Shenjian Technology.
Shan Yi, who joined this team late, is a 2004 undergraduate graduate of the Department of Electronic Engineering in Tsinghua University. In 2008, he studied for his doctorate in the Department of Electronic Engineering in Tsinghua University. Compared with Yao Song and Han Song, Shan Yi has richer experience. After graduating from his Ph.D., he joined Baidu Research Institute in 2014, joined Horizon Robot as a founding employee in 2015, and served as CTO when he joined Shenjian Technology in 2016. In October of the following year when
was established, Shenjian Technology released 6 deep learning application products. In June 2018, Shenjian announced that it would enter the ADAS field and sign contracts with major automakers in Japan, the United States and Germany.
Just one month later, Xilinx announced the full acquisition of Shenjian Technology. The amount of the transaction was not disclosed, but before the acquisition, Xilinx had completed financing of $5 million in the angel round, tens of millions of dollars in the A round and $40 million in the A+ round. The outside world evaluated that the amount of the acquisition was around $300 million.
Also started a business with the support of the mentor, there are Tsinghua Electronics Team and Paifang Technology. Wu Tongda, co-founder and CEO of Paifang Technology, is a graduate of the Sanqing University. His undergraduate, graduate student and doctoral degree are Professor Liu Yongpan, director of the Institute of Circuits and Systems, and doctoral supervisor of the Department of Electronic Engineering, Tsinghua University.
Professor Liu Yongpan's team started research on AI chips in 2014 and released the first research results of the Sticker series in 2016. In 2018, the team developed chips such as Sticker-I and Sticker-T, which achieved a huge breakthrough in the energy efficiency of chip inference.
This is also the year when Paifang Technology was established. It has passed the AI chip entrepreneurship boom in 2018 and coincides with the capital winter. However, Wutong said that Paifang Technology was not affected too much and did not join the AI chip entrepreneurship boom in 2015 because his energy was more about exploring the depth of technology with his mentor.
The other two founders of Paifang Technology also graduated from Tsinghua University. The experiences of President Ma Jun and Wu Tongda are similar and have richer industrial experience. Ma Jun is also a PhD from Tsinghua University. After graduation, he served as an algorithm expert in the Decision Intelligent Laboratory of Alibaba Damo Academy and a data scientist at China Petroleum Geophysics Company. He has 7 years of algorithm research and development experience and 4 years of industrial practice experience.
Paifang Technology Wutongda
In 2018, the two of them had the same idea and wanted to use AI technology to promote the upgrading of my country's industrial intelligence. So, the two teams came together with chip technology and algorithm technology. Wutongda is responsible for the chip team, Ma Jun is responsible for the software team, and Li Jinyang is responsible for the hardware team as the COO. He graduated from Tsinghua Electronics Department with undergraduate and master's degree, and has profound product research and development and product operation experience in monitoring systems and other projects. Just two months after the establishment of
, The Paper completed its angel round of financing. Its Sticker-I can flexibly configure the chip according to the input situation, while Sticker-T chip is a general-purpose neural network calculation based on transformation domain.With the full stack solution, Paifang Technology's AI solutions have been implemented in the field of industrial Internet equipment operation and maintenance, and are still expanding in this direction.
Tsinghua chips have two core institutions, one is the Department of Electronics of Tsinghua University, and the other is the Department of Microelectronics and Nanoelectronics of Tsinghua University (that is, the Institute of Microelectronics, which is familiar to everyone). It has similarities with the entrepreneurship of Shenjian Technology, and the entrepreneurship of Qingwei Intelligent also relies on the strong academic research of the Tsinghua team. Qingwei Intelligent's AI chip uses a technology called reconfigurable, which is a "software-defined chip" that the reconfigurable computing team led by Professor Wei Shaojun and Associate Professor Yin Shouyi has been studying for over the past decade.
Professor Wei Shaojun is currently the director of the Institute of Microelectronics of Tsinghua University and the chairman of the IC Design Branch of the China Semiconductor Industry Association. Yin Shouyi graduated from Tsinghua University with undergraduate, graduate students and doctoral degree. After finishing his postdoctoral work at Imperial University of Technology in 2007, he returned to Tsinghua Institute of Microelectronics as deputy director.
Deputy Director of the Department of Micronano Electronics, Deputy Director of the Institute of Microelectronics, and Chief Scientist of Qingweizhi, Professor Yin Shouyi
In June 2017, the first generation of artificial intelligence chip Thinker-I of the reconstructible computing research team of Tsinghua University was launched. A year later, Beijing Qingwei Intelligent Technology was established, and Professor Yin Shouyi served as the chief scientist. Ouyang Peng, another core figure of the Qingwei Intelligence team, graduated from Tsinghua Microelectronics Institute in 2014. His main research direction is "Research on key technologies for reconstructible computing for computer vision."
This entrepreneurial team with a strong academic atmosphere also has a CEO of Computer Communications at Beijing University of Posts and Telecommunications. Wang Bo not only has rich industrial experience, but also is a serial entrepreneur. Qingwei is his third entrepreneurship.
Relying on the more than ten years of technical accumulation and rich industrial experience of founding soft software, under the trend of AI, Qingwei Intelligent not only mass-produced the first reconfigurable AI voice chip in less than a year after its establishment, but also quickly used wearable smart devices, small home appliances, home appliances, toys, and in-car scenarios.
Nowadays, Qingwei Intelligent's visual AI chip industry has been released, and the implementation scenarios are also expanding.
Graduates from Tsinghua Electronics Department and Microelectronics Research Institute can also start their own businesses together. Lu Yong, CEO of Tanjing Technology, is a graduate of the Sanqing University and has a research background in the Institute of Microelectronics. After graduation, Lu Yong first worked in Silicon Valley Digital-Analog Semiconductor for 5 years, and then joined Marvell. During his 10 years at Marvell, he led the construction of a strong Marvell Chinese chip R&D team from scratch and served as the director of Chinese chip R&D. Before starting a business, he felt that he should do something in such a good entrepreneurial environment in China.
Danjing Technology CEO Lu Yong
Until 2017, after carefully studying the real needs and implementation scenarios behind the AI chip boom, Lu Yong decided to leave Marvell and founded Danjing Technology with his past colleagues, old friends and fellow Tsinghua students.
The core founding team of Tanjing Technology is about 7 people, responsible for chip design, system design, software, algorithm and other fields. Among them, Li Tongzhi, vice president of software R&D of Tanjing Technology, is a doctoral graduate of Tsinghua Electronics Department in 2003 and has rich industry experience.
This experienced entrepreneurial team in the industry has designed a storage-first architecture SFA. Unlike the common methods to solve the memory bottleneck of AI chips, SFA is a computing architecture with storage scheduling as the core. The data is calculated during the migration process between storages. Computing is just an evolution for data.
SFA architecture was originally designed for AI visual computing, but in order to be implemented faster, Tanjing Technology first implemented the chips of this architecture in the voice market. This "dimensionality reduction strike" method has enabled Tanjing's AI voice chips to ship millions in less than one year after mass production.
The two Tsinghua entrepreneurs of Hei Sesame Intelligent also made a decision and then acted. Dan Jizhang studied in the Department of Radio Electronics, Tsinghua University, majoring in Microelectronics. After graduating from the Department of Microelectronics in 1997, he joined OmniVision. 19 years later, in 2016, seeing the trend of AI vision, Shan Jizhang and his long-known old friend Liu Weihong co-founded Heisema Intelligence.Shan Jizhang was appointed CEO, and Liu Weihong was appointed COO, who was also a graduate student at Tsinghua University.
Tsinghua entrepreneurs with rich industry experience are more well known as the Horizon team. Two of the four co-founders graduated from Tsinghua University. Huang Chang, vice president of Horizon Algorithm, is a graduate of Sanqing. He has always focused on computer vision, machine learning, pattern recognition and information retrieval, and has many papers and patents. After graduating from the Department of Computer Science and Technology at Tsinghua University, Huang Chang worked as a researcher at the University of Southern California and the NEC Institute of America. In 2010, Yu Kai, founder and CEO of Horizon.
Horizon co-founder, Yu Kai, Fang Yi, Yang Ming, Huang Chang
In 2012, Huang Chang left NEC to join Baidu's American R&D Center. In 2013, he participated in the establishment of Baidu Deep Learning Research Institute (IDL), and served as a senior scientist and chief R&D architect.
From NEC to IDL, in Yu Kai's mind, Huang Chang is the unique choice for his own algorithm partners. Therefore, when Yu Kai founded Horizon in 2015, he chose Huang Chang as the vice president of Horizon's algorithm.
Another Tsinghua co-founder of the Horizon, Yang Ming, was admitted to Tsinghua University from Tianjin No. 1 Middle School. After completing his undergraduate and master's degree at Tsinghua University, he went to Northwestern University in 2004 to pursue his doctorate degree. Yang Ming, who graduated from a doctorate, worked for NEC Labs America and Facebook, and is engaged in computer vision machine learning research and product development. When he started his business with Yu Kai in 2015, he published more than 50 academic papers, cited more than 2,600 times, and obtained 13 US or international patents.
In addition to its co-founder, Xu Wei, a chief scientist of General AI, joined Horizon in 2018, also graduated from Tsinghua University and holds a master's degree from Carnegie Mellon University in the United States.
, an entrepreneurial team with both technology accumulation and rich industrial experience, has successfully mass-produced the "Journey" series processor focusing on intelligent driving and the "Sunrise" series processor focusing on AIoT, and has been widely used on a large scale.
By integrating hardware and software, Horizon accelerates the implementation of AI by providing a complete solution from chips to algorithms to cloud. In February 2019, Horizon, an AI unicorn company, received an investment of about US$600 million in Series B, with a valuation of US$3 billion.
Like the Department of University of Science and Technology of China, the Tsinghua Department also has founders who have resigned from AMD. Zhao Lidong, CEO of Suiyuan Technology, graduated from the Department of Electronics at Tsinghua University and obtained a master's degree in the Department of Electronics and Computers at Utah State University in the United States. He joined AMD in 2007. He served as senior director of the computing business department and senior director of the product engineering department. He is responsible for the research and development of CPU/GPU/APU and multiple related core IPs. The team size exceeds 1,000 people and has participated in the establishment of the China R&D Center.
Suiyuan Technology CEO Zhao Lidong
In January 2008, Zhang Yalin, who graduated from the Department of Electronic Engineering of Fudan University, joined AMD, and Zhao Lidong is his direct leader. The two worked together for about five years, and at the end of 2014, the relationship between the two colleagues came to an end. Zhao Lidong jumped to Unigroup Communications Technology Group as vice president, and Zhang Yalin remained at AMD.
Zhao Lidong, who had just joined the then Unigroup Communication Technology Group, served as the group's vice president. By March 2015, he also served as the president of Ruidike Microelectronics Company, a subsidiary of Unigroup , and was promoted to the vice president of Unigroup in March 2017.
Although they are in different companies, they have not forgotten an agreement at AMD, hoping to do some localization projects for China. After finishing AMD's last project in September 2017, Zhang Yalin chose to join the AI chip craze with Zhao Lidong.
In March 2018, Suiyuan Technology was established in Shanghai. Just 20 months later, Suiyuan Technology released its first cloud training chip, Suisi. The AI acceleration card based on this chip is said to be the highest single-precision performance in the world.
When the product was released, Zhao Lidong had already carried out close cooperation with Tencent for general artificial intelligence application scenarios. Next, he will select partners from cloud service companies, industry service companies in traditional fields, AI supercomputing centers and smart cities. Entrepreneurs with AI chips above
have made different degrees of innovation, and another entrepreneur combines AI with the rapidly developing RISC-V chips.After graduating from the electronics department of Tsinghua University, OURS CEO Tan Zhangxi chose to continue his studies abroad. He studied for a Ph.D. in computer science at the University of California, Berkeley, and studied under David Patterson, the Turing Award winner and pioneer of the RISC-V open source instruction set architecture. Influenced by his mentor, Tan Zhangxi founded OURS in 2017, focusing on RISC-V instruction-level edge AI chips.
OURS Tan Zhangxi took a photo with David Patterson
Lingxi Technology's innovation in the field of AI is also worth mentioning. Seven of the nine founders of this company are from Tsinghua University, but they are not from the Department of Electronics and Microelectronics Research Institute. Professor Shi Luping and Professor Pei Jing are from the Brain Computing Center of Tsinghua University. Professor Youhui is from the Department of Computer Science at Tsinghua University, and Professor Zhao Mingguo from the Department of Automation at Tsinghua University, Professor Song Sen from the Department of Biomedical Engineering at Tsinghua University, Dr. He Wei from Tsinghua University, and Dr. Deng Lei from Tsinghua University.
In August 2019, a paper titled "Heterogeneous Tianji Chip Architecture for Artificial General Intelligence" introduced a new type of artificial intelligence chip (combining brain-like computing and computer science-based artificial intelligence) appeared on the cover of the August issue of the top academic journal Nature, which attracted attention.
Professor of Brain Computing Center of Tsinghua University and Director of Lingxi Technology Shi Luping
The first author of this paper is Pei Jing, the supervisor of Lingxi Technology, and the corresponding author of the paper Professor Shi Luping is a director of Lingxi Technology.
The AI leader of Bitmain, a company known for its Bitcoin mining machine and its business has expanded to the AI chip field, is also from Tsinghua University. Wang Jun, who was promoted to the position of CEO of Bitmain's AI business line in January this year, was a Tsinghua graduate who was admitted to Tsinghua University in 2005 and was also the first batch of computer theory science experimental classes ("Yao Class"). After graduation, Wang Jun worked in NetEase Youdao, Google, and Baidu IDL (Institute of Deep Learning). In 2016, he joined Bitmain through a friend's introduction and became one of the earliest employees in the AI department.
Tang Weiwei, Director of AI Products at Bitmain, also has a bachelor's degree in electronic science and technology and a master's degree in automatic control from Tsinghua University.
The two major factions of China's AI chips
Obviously, most Chinese AI chip entrepreneurs have study experience at Tsinghua or University of Science and Technology of China. Interestingly, the chip leaders of Alibaba and Baidu, two domestic chip-making technology giants, are also from Tsinghua University and University of Science and Technology of China respectively.
Xie Yuan, head of the Computing Technology Laboratory of Alibaba Damo Academy, is also a graduate of the Department of Electronic Engineering at Tsinghua University who graduated in 1997. Now Xie Yuan is a leader in China in the field of chip hardware architecture, and has supported half of the architecture field with his own strength. He also achieved the Architecture Top Hall of Fame Grand Slam.
Baidu chief architect Ouyang Jian is a graduate student who graduated from the Architecture Department of University of Science and Technology of China in 2009.
In addition to Tsinghua University and University of Science and Technology, Peking University , Nanjing University , and Shanghai Fudan University have also contributed many talents to China's AI chips. Wang Shaodi, CEO of Zhicun Technology, which focuses on in-memory computing, and Guo Xinjie, CTO, are classmates from Peking University. Chen Feng, the founder of Yizhi Technology, which focuses on AI machine vision algorithms and SoC chip design, is also from Peking University.
Horizon CEO Yu Kai received his bachelor’s and master’s degrees from Nanjing University. Li Yunpeng, CEO of Tianshu Zhixin, is also an undergraduate graduate of the Department of Computer Science at Nanjing University.
Fudan University also graduated from Fudan University with the above mentioned Zhongxing Micro CTO Zhang Yundong, Yuntian Lifei CEO Chen Ning and Suiyuan Technology COO Zhang Yalin, the co-founder of Kunyun Technology also graduated from Fudan University.
Through sorting out, we can find that some of these graduates choose to start a business after graduation and have achieved good results. Many people also go abroad to continue their studies after graduating from Tsinghua University or University of Science and Technology, and choose to start a business under the influence of their mentors. Of course, many of them have accumulated a lot of practical experience in Silicon Valley and world-renowned semiconductor companies, and finally joined the AI chip army after seeing the opportunities of AI chips. Leifeng.com
In any case, we hope that under the leadership of Tsinghua University and the University of Science and Technology of China, China can cultivate more excellent AI chip talents. We also hope that more AI talents will join China's AI chip team and jointly strengthen China's AI career.Leifeng.com Leifeng.com
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-