'The industry' Ali CTO: China's chip industry is very difficult to break through | but not without opportunity

1. Ali CTO: China's chip industry is difficult to break through but it is not without opportunities; 2. Ali CTO talks about Zhongtianwei: Chip strategy has long-term commitment to long-term investment; 3. Horizon Yukai: AI chips in the United States are just beginning to avoid Iron and steel-style core

1. Ali CTO: Breakthrough in China's chip industry is difficult but not impossible;

Netease Technology News April 21, Alibaba Group Chief Technology Officer Zhang Jianfeng said in an interview with the media, Alibaba began to lay out the chip strategy four years ago. In 2014, Ali invested in Zhongtianwei for the first time, and pursued it again in 2017. A wholly-owned purchase is already a matter of course.

On April 20 of the previous day, Alibaba announced that it had wholly owned the China-based Microsystems Co., Ltd., the only independent embedded CPU IP Core company in mainland China.

It is reported that in addition to the embedded chip's Zhongtian Micro, Alibaba also has a series of investments and deployments in terms of network chips, mobile phone chips, IOT chips, etc. Up to now, Ali has also invested in the Cambrian, Barefoot Networks, Shen Jian, Kneron, ASR and other five chip companies.

At the same time, as an important part of Ali's IOT strategy, Alibaba Dharma has established a senior professional team to conduct independent research and development of AI chips.

'Alibaba's IOT strategy must be controlled by resources, technology is controllable, and it has core competitiveness. The bottom line is the chip,' Zhang Jianfeng said. 'This is also the most important research breakthrough since the establishment of Alibaba. one' .

In October 2017, at the Alibaba Cloud Conference, Zhang Jianfeng announced that Alibaba has established a global research institute under the name of 'Dharma'. Currently, it has set up scientific research institutes in multiple locations around the world, based on basic science, disruptive technologies and applied technologies. Research.

Zhang Jianfeng said that it is difficult for China to achieve breakthroughs in the chip industry, but it is not without opportunities, not to mention unattainable. Especially in certain areas facing the future, China and the United States are on the same starting line, such as artificial intelligence. He said : For example, the Internet of Things represented by urban brains and smart cities brings demand for new types of chips. At present, there are no mature chips in the global industry. This brings the opportunity for renewed competition. Fields have the opportunity to master core competencies.

'We expect a couple of years later, there will be two to three billion devices connected to the chip, and the core of the chip will be handled. The scope and scale of the chip's demand today are not at all in an order of magnitude compared to the future. It will be several hundred. Thousands of times of growth, the opportunity for China to achieve a breakthrough is in it," Zhang Jianfeng analyzed.

'This is equivalent to the appearance of another track. If breakthroughs are made in the above areas, it means that Chinese enterprises have achieved lane change and overtaking. This is also the reason why Alibaba began to lay out its chip strategy four years ago.', Jianfeng Zhang Say.

To achieve 'change lanes', Zhang Jianfeng believes that all sectors of industry, research, and research in China must work together to achieve breakthroughs. Because the 'chip strategy is not an eternal task, enterprises including Alibaba are determined to invest for a long time. They must also have a future-oriented strategy. Vision and tolerance of loneliness perseverance'.

Finally, Zhang Jianfeng concluded that in the field of chips, as in any other field of the era of globalization, there is both global competition and global cooperation. In the global market, there is no one else, and this is the basis of reality. It is also China's production, research, and research. The direction of joint efforts of all walks of life. (Peng Lihui)

2. Ali CTO talks about Zhongtian Micro: The chip strategy is non-permanent. It requires long-term investment;

Alibaba established Daruma House and CTO Zhang Jianfeng made a speech

Sina Technology News on the evening of April 21, after Alibaba’s wholly-owned acquisition of Zhongtian Micro Corporation, Alibaba Group’s chief technology officer Zhang Jianfeng disclosed for the first time the company’s long-term layout and planning in terms of chip strategy. Zhang Jianfeng stated that the company was back in the fourth quarter. Year ago, it began to lay out the chip strategy. 'In 2014, Ali invested in Zhongtianwei for the first time. In 2017, he pursued another investment, and this year's wholly-owned acquisition is already a matter of course.

Zhang Jianfeng stated that in addition to the embedded chip's Zhongtian Micro, in terms of network chips, mobile phone chips, IOT chips, etc., Alibaba also has a series of investment and deployment. As of now, Ali has also invested in the Cambrian, Barefoot Networks. , Kneron, ASR, and other five chip companies. At the same time, as an important part of Alibaba's IOT strategy, Alibaba Dharma has established a senior professional team to conduct independent research and development of AI chips.

'Alibaba's IOT strategy must be controlled by resources, technology is controllable, and it has core competitiveness. The bottom line is the chip,' Zhang Jianfeng said. 'This is also the most important research breakthrough since the establishment of Alibaba. one' .

Zhang Jianfeng pointed out that it is very difficult for China to achieve a breakthrough in the chip industry, but it is not without opportunities, not to mention being unattainable. Especially in certain areas facing the future, China and the United States are on the same starting line, such as artificial intelligence. He said For example, the Internet of Things represented by urban brains and smart cities brings demand for new types of chips. At present, there are no mature chips in the global industry. This brings the opportunity for renewed competition. Fields have the opportunity to master core competencies.

'We expect a couple of years later, there will be two to three billion devices connected to the chip, and the core of the chip will be handled. The scope and scale of the chip's demand today are not at an order of magnitude compared to the future. It will be several hundred. Thousands of times of growth, the opportunity for China to achieve a breakthrough is in it,” Zhang Jianfeng analyzed. “This is equivalent to the appearance of another track. If there is a breakthrough in the above areas, it means that Chinese enterprises have achieved a change of course. Overtaking. This is why Alibaba started to lay out its chip strategy four years ago.'

Zhang Jianfeng believes that all industries, universities, research institutes, and universities in China must work together to achieve breakthroughs. Because the chip strategy is not an overnight task, enterprises including Alibaba are determined to invest for a long time, and they must also have a vision for the future and be able to withstand the loneliness. Persistence'. In the field of chips, as in any other field in the era of globalization, there is global competition and global cooperation. It is also the basis of reality in the global market, and it is also the basis of all industries in China. The direction of efforts. (Dapeng)

3. Horizon Yu Kai: The AI ​​chip in the United States has just begun

Horizon Yu Kai: On the AI ​​chip runway, China and the United States are just beginning

The 'core pain' brought by the 'ZTE Incident' has lasted nearly a week.

The U.S. Department of Commerce announced on April 15 that the ban on U.S. companies selling electronic components, equipment and software technology to ZTE within seven years had triggered concerns about the core competitiveness of the Chinese chip industry.

At the same time, Ali first announced that it was developing a super AI chip, and then announced a wholly-owned technology company to acquire a large-scale production chip.

Will domestic AI chips solve China's 'core pain'? What role will AI chips play in the future development of Chinese chips?

Dr. Yu Kai, founder and CEO of AI Chip Enterprise Horizon, a specially-received national expert from the Ministry of Organization's 'Thousand People' Project, told Sina Finance that he believes that the 'Zhongxing Incident' has further promoted the AI ​​chip to become a public concern for all parties in the industry. But also calm, avoid all people make the core '.

The AI ​​chip in the United States is just beginning, and the opportunity is still great

According to the "Twenty Group National Innovation and Competitiveness Yellow Paper" issued by the Development Research Center of the State Council, China's key core technologies are highly dependent on the outside world, and 80% of chips rely on imports. Some experts pointed out that China's chip technology and production processes are currently on the whole. The United States cannot compare, but it can be completely self-sufficient in the production of low-end chips. The problems are mainly in the industries and technologies of high-end products.

Yu Kai said that high-end chips are still in a follow-up phase. This is an objective fact. 'The chip's design, R&D, production, test and verification to the final scale application is a very complicated system engineering, with long cycle, large investment, and risks. High, therefore, requires long-term accumulation of technical experience. It cannot be accomplished overnight. Foreign countries have been plowing for many years in this field, but China has started late.

For the differences between China and the United States in manufacturing chips, Founder and CEO of OURS Foundry and CEO Tan Zhangyi of Silicon Valley's 'Returnee' AI Chip Company stated frankly that apart from hardware differences, there are also differences in the level of software knowledge. 'In addition to the chip itself, many The technology for making chips is also in the United States'.

However, AI chips are different from conventional chips. Yu Kai believes that 'AI chips are scenarios, data, algorithm-driven, and require the coordination of algorithms and hardware design. In this track, everyone (Sino-US) is actually on the same starting line. It's just beginning, so the chance is great. '

For the current AI chip industry, Dr. Zhangzhang Zhang believes that with the emergence of big data, for enterprises that need data processing, the application level AI chip is just needed.

Just one day before Ali announced its own research and development of AI chips, Facebook's engineer recruitment notice confirmed that the US Internet giant will also enter the AI ​​chip.

Dr. Yu Kai said that currently China and the United States have become leaders in international artificial intelligence technologies and industries, and artificial intelligence chips have become the focus of focus. Although we have a large gap with the United States in the field of traditional integrated circuits, the characteristics of artificial intelligence algorithms have brought new The evolution of the computing architecture and ecosystem has brought us an opportunity for development. 'The era of smart interconnection of all things has come, and the number of smart terminals that we will access will be trillions, and the scale of artificial intelligence chips driven by it will be far greater. In the Internet era, artificial intelligence chips are also considered to be the commanding heights of artificial intelligence technology and industry development in the future.

China AI chip pattern: Powerful automatic driving processor and other fields

What is the development of domestic AI chips?

At present, the head companies of domestic AI chips include the Horizon, Cambrian, and Shenzhen Science and Technology. The Horizon completed more than US$100 million in Series A financing in December last year. The company is currently conducting Series B financing; The new round of funding for the Cambrian this month has been carry out.

Each at the same time competes for different application areas, such as the current AI chip independently developed by Ali, mainly to solve the AI ​​reasoning and calculation problems of commercial scenes such as images, video recognition, and cloud computing, so as to improve the operation efficiency and reduce the cost.

Dr. Yu Kai introduced the first generation of the world’s leading embedded AI vision processor currently being developed by Horizon. It is mainly used in areas such as autopilot, smart security, smart city, smart business, etc. The chip mainly solves problems such as multi-sensor fusion of autopilot processor and environment perception modeling.

Yu Kai said earlier that the horizon will force the development of automatic driving processor, 'I hope to continue driving on a very long runway like autopilot. By 2025, 30 million vehicles will be loaded with China's own chips, making us really This area has its own right to speak. '

'When self-improvement, but avoid steel-type people's core

Regarding the concern raised by the 'ZTE Incident', Dr. Yu Kai believes that the high-tech industry represented by the electronic information industry in China does indeed have a weak foundation in key chips and core components. However, Yu Kai still expressed optimism that 'chips as information' The core and cornerstone of industrial development requires more efforts to tackle problems. In recent years, integrated circuits have also become China's national strategy and have increased investment and layout. We believe that in the future, the country will further strengthen the top-level design, which will promote China's chip industry and even the entire industry. The development of the chain is very favorable.

The key three factors that promote the development of artificial intelligence are algorithms, data and computational power. Yu Kai stated that the chip that carries the computational power is the cornerstone of the development of artificial intelligence and is the most important infrastructure resource, especially on the terminal side. Consumption, performance, and stability all have very high demands. Of course, this incident has further promoted the AI ​​chip to become a enthusiasm for all parties in the industry. However, we must also clearly understand that chip R&D is a complicated project that requires a scientific and rigorous attitude. We should be self-reliant. , But also calm, avoid steel-based people made core. Sina Finance

2016 GoodChinaBrand | ICP: 12011751 | China Exports