On September 18, the 4th Huawei Full Connectivity Conference opened in Shanghai. Huawei released its computing strategy for the first time, based on architecture innovation, investment in the full scene processor family, business strategy of doing something and not doing something, and building an open ecosystem. This opened a new era of navigation for the computing industry. At the same time, Huawei launched Atlas 900, the world's fastest AI training cluster, to accelerate the intelligent process of scientific research and business innovation.
Hu Houkun, Vice Chairman of Huawei, delivered a keynote speech
The main points of Hu Houkun's speech are as follows
Two years ago, Huawei released a new vision and mission, which is to "bring the digital world into every person, every family, and every organization, and build an intelligent world of interconnected things".
We believe that the intelligent world has three characteristics, namely, everything perception, everything interconnection, and everything intelligence. To support such an intelligent world, there are two key technologies that need continuous innovation and investment: connectivity and computing.
Two key technologies of intelligent world
Huawei has always been known as a connectivity company. In the past three decades, Huawei has made continuous investments in the connectivity field. From wired to wireless, from 2G, 3G, 4G to today's popular 5G, Huawei has made some achievements. However, Huawei is not just a connection company. In our view, to support the future intelligent world, both connectivity and computing are indispensable. In fact, the two technologies of connection and computing are like twin brothers, promoting and developing together. It can be said that in the future intelligent world, where we connect, computing will go, and where there is computing, there will be connectivity. In the past, we talked a lot about connection. In fact, Huawei has invested in computing for ten years. Today, I would like to focus on sharing Huawei's insights and strategies in computing. Computing becomes an extension of human ability First, let's talk about our observations on the computing industry. Since the birth of the first computer in 1946, more than 70 years ago, from mainframe to personal computer, from desktop to laptop, to tablet, from smart phone to wearable device, computing devices have become smaller and smaller in size, more powerful in function, and closer to people. It can be said that computing power has become an extension of human power. We believe that the development prospect of computing industry is limitless along this trend. The computing model continues to evolve Over the past 70 years, the computing model itself has also been developing. The first time I came into contact with computers was in college. At that time, when I learned programming, my teacher taught me that if you give me a formula, it will give you a result. In fact, this is a rule-based computing model. Such a calculation model has solved the problems of population census, ballistic calculation and so on.
However, for some problems that can not be described by certain rules, such as speech recognition, image recognition, real-time translation, etc., rule-based computing mode is difficult to solve these problems.
Fortunately, scientists have found a new computing model, the statistical computing model, which can solve these problems mentioned above. Moreover, this statistical based computing model has also become the key cornerstone of the current development of artificial intelligence.
We believe that this statistical based computing model will gradually become the mainstream. It is estimated that in five years, the computing power consumed by AI computing will account for more than 80% of the total computing power consumed by the whole society. Based on such changes, we believe that computing is entering a new era of intelligence. Computing enters the era of intelligence In the era of computing intelligence, we believe that there are three important characteristics. The first feature requires strong computing power. Statistical computing itself is a kind of violent computing, highly dependent on computing power. For example, in order to let the computer know a cat, it needs millions of pictures to train, which is very surprising for the consumption of computing power. For more complex scenes such as automatic driving, astronomical exploration, weather prediction, the demand for computing power will be greater. The second feature, computing and intelligence, will be everywhere, not just in the center. From violence computing at the central node to professional computing at the edge, such as gene sequencing, and personality computing at the end, such as earphones and mobile phones, together constitute the computing form of the future intelligent era. The third feature is the need for efficient collaboration between end cloud and edge cloud. The center side is responsible for the calculation of the general model, providing collaborative support for the personalized calculation at the end side and the professional calculation at the edge side. 2 trillion yuan of computing industry To achieve these three characteristics, we still face many challenges. For example, super computing power, exploration of new computing architecture, full scene oriented processors, etc. In Huawei's view, challenges just mean opportunities. The greater the challenges, the greater the opportunities. We are facing a big blue ocean of computing industry. Citing Gartner's data, it is estimated that by 2023, the scale of computing industry will exceed $2 trillion.
Huawei's Computing Industry Strategy
Therefore, Huawei is firmly committed to the computing industry. We mainly focus on four aspects, including breakthroughs in architecture innovation, investment in the full scene processor family, adhering to the business strategy of doing something and doing nothing, and sparing no effort to build an open ecosystem.
From our previous insights, computing is ubiquitous and intelligent in the future, in which computing power is the key foundation.
At present, the supply of computing power in the industry is scarce. The key to the supply of computing power is the efficiency of processors. At the moment when Moore's Law has almost reached its limit, from the perspective of industrial development, we must seek breakthroughs in processor architecture and use new processor architecture to match the growth of computing power. On the other hand, from the perspective of Huawei's own business layout, there are both network services and terminal services, as well as public cloud services, which comprehensively cover the end, edge, and cloud. Therefore, full scenario intelligence itself is the basic requirement for Huawei's business intelligence. Therefore, we have launched the Da Vinci architecture to achieve ubiquitous computing and intelligence, which is a natural choice. So far, the Da Vinci computing architecture is the only processor architecture in the industry that can cover the whole scenario of "end, edge and cloud", which is also a solid foundation for Huawei to build the computing industry.
Investment in full scenario processor
Processors are the most fundamental part of the entire computing industry. After years of investment efforts, Huawei has released several series of processors.
Specifically, it includes Kunpeng series supporting general computing, Shengteng series supporting AI, Kirin series supporting intelligent terminals, and Honghu series supporting smart screens. In the future, we will continue to invest in processors, and we will also launch a series of processors for more scenarios. Business strategy "do something and do nothing"
Our business strategy, in a word, is "doing something and not doing something". In general, we do not directly sell processors to the outside world. We focus on cloud services for customers and components for partners. We give priority to supporting partners to develop complete machines. Specifically, we have three considerations:
Hardware opening: We open the server motherboard, AI module, board card and other hardware to partners to help them do a good job of the whole machine and solutions.
Open source software: We open source the server operating system, database, AI development framework and other software to help partners do a good job of commercial version and make software development easier.
Enable application development and migration: We don't do applications, but we provide tools and teams to help partners develop and migrate applications more efficiently.
We hope to better support our partners through such a business strategy of "doing something, doing nothing". Finally, the computing industry is an open industry, highly dependent on ecology, and requires global cooperation. In 2015, Huawei released the fertile land plan for the first time. In the past four years, we have made very smooth progress, and have developed more than 1.3 million developers and more than 14000 ISVs. Today, we will upgrade the fertile land plan, continue to invest 1.5 billion dollars, expand the scale of developers to 5 million people, and enable global partners to develop applications and solutions. On the third day of this conference, my colleagues will release a detailed plan. General Computing Business Strategy In the field of general computing, we invest in the Kunpeng series to build a competitive general computing processor.
On this basis, we are also continuously investing in key technologies and products such as boards, servers, operating systems, databases, compilers, etc., hoping to open up the whole ecological chain through strong investment, so that our partners will be more confident and more determined to participate in the development of Kunpeng industry. We are also working with our partners to build Kunpeng Industrial Ecological Base.
We work with local governments and partners to build an innovation base and incubation platform for Kunpeng computing industry in combination with local advantages. Take the platform as the carrier, gather ecological partners, carry out application demonstration, cultivate industrial talents, and incubate industrial standards.
At present, the work is going very smoothly. We have started to implement the implementation in Beijing, Shanghai, Shenzhen and other cities. It is expected that more partners will join us in the future. AI calculation: Full stack and full scenario AI solutions are fully implemented
Last year, at the Full Connect Conference, Huawei released an AI solution for the whole stack and scenario. To be frank, we only delivered the Shengteng 310 processor and the ModelArts application development platform for reasoning. This year, our Ascension processor and AI computing framework MindSpot for training have been released. So far, Huawei's full stack and full scenario AI solutions have been fully implemented, and you can be assured to cooperate with us.
Release Atlas 900: The fastest AI training cluster in the world Today, we will release a heavyweight product - Atlas 900. This product gathers Huawei's decades of technological precipitation and is the world's fastest AI training cluster, consisting of thousands of Ascension processors.
In the gold standard ResNet-50 model training, which measures AI computing power, Atlas 900 completed the training in 59.8 seconds, 10 seconds faster than the original world record. What is this concept? Quite Finish at the sprint champion At the end of the journey, after drinking a bottle of water, we waited for the second place. We believe that the powerful computing power of Atlas 900 can be widely used in scientific research and business innovation. For example, astronomical exploration, oil exploration and other fields require huge data calculation and processing. It used to take months of work, but now it is just a matter of seconds to hand over to Atlas 900. Next, let's share a case of astronomical exploration, which was jointly created by Huawei, the Shanghai Astronomical Observatory and SKA. Astronomical research is highly dependent on massive data analysis and calculation, and cannot be separated from strong computational support. The following is a map of the stars in the southern hemisphere. There are 200000 stars on this map, which cannot be seen with the human eye. These images are from the data of SKA radio telescope. In order to enable you to use the super computing power of Atlas 900 immediately, we have deployed it to Huawei Cloud and opened it to global scientific research institutions and universities at a very favorable price. You are welcome to apply.
Open up a new era of computing industry navigation Today, the digital and intelligent transformation of all walks of life is in full swing, and the computing industry is booming. Huawei is firmly committed to the computing industry, starting with the most difficult architecture breakthrough, and developing its own processors. Based on this, Huawei builds an industry and an ecosystem.
We firmly believe that only by overcoming the greatest difficulties can we truly build competitiveness and open up greater industrial space for industry partners. We should use the best technology to solve the world's most difficult problems and make the impossible possible. This is a new era of great navigation in the computing industry. What we are looking forward to is the competition of thousands of sails, not the striving of a single boat. We look forward to working with partners from all walks of life to seize the historical opportunities and create a new height of intelligence.
