Adhere to globalization and significantly increase strategic investment in root technologies to continuously enhance medium- and long-term competitiveness
In the economic sphere, an action, a custom, an institution or a law can have not only one consequence, but a series of consequences. Of these consequences, only the first occurs immediately, it is revealed in sync with the cause, it is visible; the other consequences are presented one after another, they are invisible. If we can foresee them, that is advantageous for us.
—Frederick Bassia, Visible and Invisible (1850)
Reviewing huawei's corporate actions and external public opinion feedback in the three years since huawei was sanctioned by the US government into the winter dormant period, it will find such a phenomenon: people pay attention to the explosive information that the company initially has a certain amount of explosive information when it is in crisis, but will ignore the long process of the company's difficult and complex self-help in the long-term process and its strategy behind it.
Unless it's an insider.
In September last year, at an internal symposium with Huawei researchers, Ren Was asked a question: "How to balance the two sides of life" and "have a future"? His answer roughly consisted of two meanings.
First, in the talent system, Huawei's scientific researchers will divide the soldiers into two roads, one way to take the road of thorough scientific research, "breaking the sky"; the other way is to "throw pen from rong" and use the basic theories mastered to solve practical business problems.
Second, only when all Huawei employees understand the company's strategy will the company have strength.
Note that this is actually a talent strategy. Within a company, a talent strategy typically serves the overall business strategy.
Huawei has demonstrated its Business Continuity Management (BCM) capabilities to deal with short-term crises over three years; balancing and differentiating between business segments in a dramatically changing and unfavorable external environment.
Some of these cases have been written into the practice cases of business schools, the way enterprises spend the winter is very different, and the efficiency and determination of stress response and strategic relocation layout are usually the core indicators that determine whether enterprises can successfully survive the life and death barrier.
However, Huawei's long-term approach to wintering actually jumps out of the short-term wintering strategy, which is a hidden proposition based on long-term development.
On March 1, during MWC22 Barcelona, Guo Ping, Huawei's rotating chairman, delivered a keynote speech entitled "Upward, Light up the future" online. In this speech, Guo Ping answered the core strategy hidden behind it.
Guo Ping said that Huawei will adhere to globalization, significantly increase strategic investment in root technologies, and strive to achieve three reconstructions of basic theory, architecture and software, so as to continuously enhance Huawei's medium- and long-term competitiveness.
How to do the "root technology"? What is the relationship between the three refactorings of basic theory, architecture and software? What is its qualitative impact on Huawei's medium- and long-term competitiveness? What kind of path will Huawei embark on? These are the questions this article is trying to answer.
The importance of root technology
Before 2019, Huawei's R&D innovation was more of an engineering innovation. The connotation of engineering innovation can be understood as the innovation of products and solutions. At the heart of engineering innovation is helping customers and partners to become more competitive, increase revenue or reduce costs, and achieve business success. In the past, Huawei had a lot of engineering and technological innovation in the fields of wireless, optical networks and smartphones.
After 2019, Huawei's strategy in the field of technology is to go one step further on the basis of engineering innovation, and one step forward in the breakthrough of basic theory and the invention of basic technology.
In today's global business system, the innovation of basic technologies is usually done by scientific research institutions and academic institutions, and the focus of commercial companies is not on technical research itself, but on the "commercialization of basic technologies". At the internal symposium mentioned above, Ren Zhengfei mentioned that basic research driven by curiosity and applied research driven by business value in basic innovation may also be combined to create both scientific knowledge and business value.
This is the "Pasteur Quadrant" innovation advocated by Professor Stokes of Princeton University in the 1990s, and the main reason why the new Endless Frontier Act proposes to transform the American Science Foundation into a science and technology foundation in 2020.
In his March 1 speech, Guo Ping mentioned that to see the future, you must look up, not be blinded by political and factional propaganda. Huawei must maintain its competitiveness in the medium and long term, look up, look forward, and find a way forward.
What is the way forward? Guo Ping's judgment is that digitalization and carbon neutrality are neutral. He said: These are two major issues in today's world.
The important strategic value of the digital economy in the global economic development has been recognized. Some forecasts say that this year, the digital economy will account for more than 50% of global GDP, which is a decisive turning point in global economic development. But the technical theories underpinning the development of the digital economy, Shannon's theorem and Von Neumann's architecture have encountered bottlenecks. Guo Ping said that this is the basis for judging huawei's progress towards basic scientific research.
The other way forward is carbon neutrality. Carbon neutrality poses challenges for the ICT industry, and Huawei evaluated the Deep Earth project, which alone monitors global environmental data to emit carbon equivalent to 12.5% of current global emissions. Further developments in digitalization will lead to increased carbon emissions, but ICTs can help other industries mitigate their "carbon footprint", which is more than 10 times their own "carbon footprint" in order of magnitude.
Guo Ping said that the density of the connection multiplied by the precision of the calculation is the strength of the digital economy. But now, Huawei not only needs strength, but also long-term vitality. Therefore, we need to consider one more dimension, that is, the strength of carbon emission reduction.
In order to achieve this goal, Huawei will significantly increase its strategic investment in "root technology" and strive to reconstruct the technical base.
The so-called "root technology" are those that can derive and support one or more technical clusters. Root technology is the root of the technology tree, which continues to provide nourishment for the entire technology tree, and to a large extent determines the prosperity and decline of the technology tree.
Three characteristics of root technology cannot be ignored.
First, the concealment, usually the root technology is difficult to be distinguished, and the root technology in some fields is often entangled.
The second is proliferation, root technology with strong vitality, once a new breakthrough is achieved, the entire technology tree may be renewed.
The third is abundance, root technology contains a lot of silent knowledge, which is often difficult to obtain through formal dissemination, but needs to be continuously accumulated through "dry learning".
Today, with the rapid advancement of the new round of scientific and technological revolution and industrial change, who can take the lead in developing new root technologies and forming a root industry that dominates the future economy may become the leader, shaper and leader of the future global economic development.
Judging from the experience of the technological development cycle in the past few hundred years, once the root technology is discovered and developed, it will "break through the ground" in a short period of time and grow a root dry industry, which has the characteristics of high added value, strong dominance over other derivative industries, and sufficient risk resistance and resilience.
On the contrary, if you cannot have a sufficient number of root technology and root industry, a country's economic development and scientific and technological innovation will inevitably encounter a ceiling and be stuck in the neck. Companies at the forefront of the industry also face such problems.
Mastering the initiative of the next generation of root technology and root industry is very important for a country and an enterprise.
Three refactorings
Guo Ping proposed that Huawei is working hard to achieve three reconstructions: basic theory, architecture, and software. These three restructurings will support the long-term sustainable development of the ICT industry.
Theoretical reconstruction is the foundation of the foundation, and Guo Ping mentioned channel capacity increase.
Ren Zhengfei proposed in 2016 that the limits of Shannon's theorem and Moore's law, and the theory of large flow and low latency have not yet been created, "Huawei has felt that the future is uncertain and cannot find a direction." Huawei has advanced in the trek. In other words, there is no progress in basic science research, and Huawei's application layer technology research and development is almost to the ceiling."
Shannon's theorem can be simply understood as, if there is noise in the channel transmitting information, the transmission of information can also be realized, and then the maximum effective rate of information transmission can be calculated according to parameters such as bandwidth. At its core is the signal-to-noise ratio. Both channel capacity and channel bandwidth are extensive, and only the signal-to-noise ratio truly reflects the technical difficulty.
Guo Ping said that the channel capacity is close to the ceiling. Huawei continues to explore theories and technologies such as a new generation of MIMO and wireless AI to further approach the Shannon limit, and at the same time research new theories such as semantic communication, trying to surpass the Shannon limit and open up a broader development space for communication.
Huawei's new generation of basic research and development is further approaching the Shannon limit Chart source: Huawei
In simple terms, an important basic research work huawei has done in these years is to redefine the new architecture of wireless networks in the post-Shannon era.
Moreover, Huawei has broken down this seemingly advanced work into ten more specific research topics. They include challenging issues such as privacy protection, information modeling, channel capacity, approximate computing, and optimization, in the form of self-research + joint efforts with the world's top scientists.
The new research results include: the theoretical model of trusted network in the AI era, the dynamic traffic prediction model of network, and the theoretical model of distributed dynamic sharded network are proposed, which greatly improves the robustness of the network, improves the predictability and network utilization of the network. Active control network system efficiency increased by 56%. At the same time, a new breakthrough has been achieved in the core theory and scheduling algorithm of the next generation of Massive MIMO, increasing the edge user experience rate by more than 30%.
The latest news is that Huawei has broken through the Shannon limit to a certain extent through the productization of Massive MIMO, and is still actively exploring the theory of distributed MIMO to provide basic theoretical support for 5G and the next evolution.
The second is architectural refactoring.
Wireless communication still faces major technical challenges such as high frequency, ultra-large bandwidth, and ultra-high speed, which is a common problem faced by the information and communication industry. In addition to trying to break through the path in theory, Huawei has refactored architecture in another direction of exploration in recent years. Guo Ping mentioned two examples. One is wireless + light; one is a new computing architecture based on AI.
Guo Ping said that the introduction of photoelectric fusion technology to solve key problems is a technical path to break through the process bottlenecks faced by future chips. It should be noted here that Huawei is a global technology leader in the field of wireless communication technology and optical technology.
Wireless communication is the starting point of Huawei, and it is also the strongest area of Huawei technology research and development. The world has entered the 5G era, and Huawei's 6G pre-research has begun, starting from the breakthrough of basic theory, researching new air interface technology, new network architecture, and other key technologies.
Let's take another example of the light field. Huawei is the world leader in optical wavelength division technology.
Taking the 400G capacity and performance of optical communications as an example, Huawei provides public data showing that at the same distance, the capacity is 40% more than the industry; under the same capacity, the distance is 20% more than the industry. These two sets of data benefit from Huawei's super oDSP algorithm.
Mayevsky, a mathematician from Huawei Russia, was the chief developer of the 5G Polar code in 2017, and now has achieved a major breakthrough in the 400G transmission algorithm, infinitely approaching the Shannon limit, leading the industry for more than 3 years.
The latest news is that in the second half of this year, Huawei will launch pluggable 400G modules, with unchanged performance, 70% reduction in size, and 50% reduction in power consumption. This module features the integration of five separate devices into a fingernail-cap-sized COSA device.
But the fusion of these two technologies and the realization of the effect of 1+1>2 are actually entering the no-man's land of new technologies.
Huawei Wireless + Optical Fusion Technology Source/Huawei
The other is a new AI-based computing architecture.
The current contradiction of computing architecture is that AI and big data applications are booming, while traditional computing architectures are still "CPU-centric". Huawei's architectural innovation idea is to design a "peer-to-peer" architecture so that GPUs, NPUs, etc. can better support business development.
Huawei's Next-Generation Computing Architecture Evolution Roadmap Chart Source: Huawei
For example, in the field of deep learning, jump out of the multiplicative neural network framework, through collaborative design with hardware, for the first time use addition to build a high-precision visual neural network, obtain more than ten times the energy consumption and chip area benefits, and open source the test code of the additive neural network, and jointly explore the new route of the next generation of artificial intelligence acceleration with the academic community.
The third refactoring mentioned by Guo Ping is software refactoring.
When it comes to software, it is inevitable to mention artificial intelligence. The direct result of the explosion of artificial intelligence technology and applications is a sharp increase in the demand for computing power, but the reality is that the hardware process that provides computing power is slowing down.
There are two solutions, one is to improve the hardware, and the other is to improve the software. Huawei has dabbled in both paths, but the results on the second path are more obvious. For example, through Hongmeng and Euler, the computing power potential of diversified hardware is more effectively utilized; through the Mindspore framework, it helps scientists and engineers improve development efficiency.
Huawei's AI-centric full-stack software reconstruction is expected to create a new ecosystem and bring new opportunities to customers and the software industry.
Guo Ping also mentioned that in the 2021 EU industrial research and development investment scoreboard, Huawei ranked second. He said that Huawei is increasing investment in system architecture optimization and doubling software performance. Because only products with full coordination of software and hardware can really bring a good user experience.
He said that the continuous investment in root technology requires patience, and it will gradually be reflected in the competitiveness of products.
In the most difficult times, how to do overseas markets?
Returning to the beginning of the question, an important wintering logic of Huawei in the past few years is that in addition to using various ways to quench the "near thirst", it has also strengthened the layout of "root technology" from the source to ensure the "far water" of future development, which fundamentally ensures its medium- and long-term competitiveness.
Huawei is a highly globalized enterprise, after being sanctioned by the US government, Huawei has encountered multiple problems in overseas markets, the pace of development has slowed down significantly, and the domestic market has become a ballast stone. During the period, Guo Ping was repeatedly asked a question by people from different fields: Will Huawei gradually withdraw from overseas markets?
On March 1, Guo Ping publicly responded to the question: "No." He said that Huawei will remain an active member of the world and unswervingly implement its globalization strategy in terms of standards, talent and supply chain.
At this time, it is still necessary to gain a firm foothold in the international market, and the only thing to rely on is excellent technical strength that can be confirmed in the short term.
As mentioned above, once the root technology is discovered and developed, it will "break through the soil" in a short period of time and grow a root industry, which has the characteristics of high added value, strong dominance over other derivative industries, and sufficient risk resistance and resilience.
In 2021, entering the third year of sanctions imposed by the US government, a cooperation project between Huawei and Ireland K2 is representative. K2 is a multinational well-known IDC operator (Colocation), in 2021 in Dublin, Ireland 16MW, 1980 cabinet cloud data center project, the first time to use Huawei intelligent indirect evaporative cooling solution (IEC), can directly use Dublin's perennial low temperature resources and water evaporation heat dissipation, to achieve year-round natural cooling, to achieve 1.15 PUE, 16MW data center annual power savings of more than 7 million kWh.
Huawei's IEC solution, one box and one system, reduces the four heat exchanges of traditional large data centers to one, simplifying the hot swap link.
Thanks to the outstanding performance of Huawei technology, the next phase of K2 Dublin is also 16MW, and the data center of the 1980 rack also adopts a smart IEC solution, which together is expected to reduce electricity consumption by 14 million kWh per year and reduce carbon emissions by nearly 70,000 tons in 10 years.
Huawei's data center deployed in k2 Source/Huawei
During Telecom Barcelona in early March, Brazilian operator TIM signed a Memorandum of Understanding on Joint Innovation cooperation with Huawei on 5G City.
Brazilian telecom operator TIM Participacoes SA is a subsidiary of Telecom Italia (TIM S.p.A.) in Brazil. TIM and Huawei have already partnered to deploy 2G, 3G, 4G and 5G networks in Brazil.
The cooperation between the two sides this time is to promote the construction of a smart city in Brazil through the 5G city model project. The first city to cooperate will be the Brazilian green city of Curitiba.
There are two deep meanings behind this cooperation.
First, the cooperation is not a wireless network construction in the traditional sense, but a cooperation of Huawei's smart city project based on 5G networks, which means that Huawei's government and enterprise digital business is an important node in Latin America.
Second, whether it is Huawei or TIM, they have high hopes for the construction of the first city, Curitiba, hoping that through its sample effect, it can radiate to more countries and regions around the world.
Huawei went abroad in 1996 and spent two decades step by step from Southeast Asia, Africa, Latin America, and the Middle East to Europe and North America. In 2019, the first year of sanctions by the US government, Huawei's overseas sales revenue was 352.1 billion yuan, compared with 306.7 billion yuan in 2020. At the end of December last year, Guo Ping mentioned in his New Year's speech that Huawei's sales revenue in 2021 is expected to be about 634 billion yuan, down 28.9% from last year. Guo Ping did not disclose more detailed regional data, but further contraction of overseas markets is inevitable.
At a time when the international environment is delicate and unchanging, it is not easy for Huawei to gain a firm foothold in the international market.
Huawei's revenue share in 2019 (left) and 2021 (right) around the world Figure/ Huawei annual report
Huawei, which has been running wild for more than 20 years, has once again returned to the starting point of the problem, but this time the problem is more practical. What kind of rhythm is most appropriate between the near thirst of the moment and the "far water" interaction it looks up to? How will Huawei's current actions and visions be connected?
At the very beginning of his speech on March 1, Guo Ping used a movie plot called "Don't Look Up" as a metaphor for the world facing a major disaster, but some people would rather turn a blind eye to it.
"This is not our future. To see the future, we must look up, look ahead, and find the way forward. He said. (The author of this article is Xie Lirong)
There's so much more to see