Huawei Global Analyst Summit 2021

Hello and good afternoon to everyone here with us in Shenzhen, and good afternoon or good morning depending on where you are in the world to those of you joining us virtually. Welcome to the 18th Huawei Global Analyst Summit under the theme 'Building a Fully Connected, Intelligent World.' My name is Simon Webb and I'll be moderating the coming sessions. We are proud to bring together many hundreds of guests from industry, analyst firms, the domestic and international media, as well as our industry partners. We only need your attention for a mere two hours today, but we have a lot to cover in that time. We have two speakers: first, Mr. William Xu, Huawei Director of the Board and President of the Institute of Strategic Research. He will be followed by Mr. Eric Xu, Huawei Deputy Chairman and Rotating Chairman. And we will conclude with a question and answer session. So without further ado, let's get started. To take us through how we will stride towards the intelligent world of the year 2030, please welcome to the stage Mr. William Xu.

Ladies and gentlemen, good afternoon, including our friends joining online. Hello everyone. Last year, a friend of mine told me that he has been joining our HAS summit for 15 years in a row. I hope that he's still here with us today. And this again is our 18th Huawei Global Analyst Summit. We have all had a difficult year facing many new challenges because of the pandemic and anti-globalization. Today, we are at the beginning of a new decade that will bring new uncertainties and opportunities. The ICT industry will also face new challenges and needs to make new breakthroughs. Population and energy underpin the well-being of our society. A UN report shows that by 2030, the global population will reach 8.6 billion, where over 12 percent will be 65 or older. Can we define a person above the age of 65 as an elderly person? Well, we have a question mark there, but still, this is a quite serious set of statistics. As for the percentage of people aged 25 or younger, it is also decreasing. Therefore, aging populations and labor shortages are hindering our social progress. There is an increasing need for solutions that not only prolong life but also improve quality of life and ensure dignity during end-of-life care. Global energy consumption is growing at an annual rate of 1.7 percent. Energy consumption has increased by 22 times since the 18th century according to a report, and currently 85 percent of the energy comes from fossil fuels. Energy sustainability is a daunting universal challenge. Therefore, low-carbon energy, electricity, and intelligent transformation are needed to ensure sustainability. We predict that by 2030, more than 50 percent of energy will come from renewable sources, more than 50 percent of cars sold will be electric, as mobility via electricity will become the mainstream. AI will change everything as more than 80 percent of homes will have smart robots. What kind of role can the ICT industry play? Well, if these numbers pan out, ICT technologies could help reduce global carbon emissions by 20 percent over the next decade by empowering other industries. This is quite a challenging goal. We have many expectations for the future that will require us to transcend our limits. We hope we can exceed the biological limits of perception. Now, our phone cameras have achieved 100 times zoom thanks to the digital technology, but there's still a huge gap between camera capabilities and what animals can achieve in nature. You may not like spiders, but spiders are much better at detecting the outlines of objects and motion than mankind. So is it possible that we can learn from them to leverage their capabilities so that we are able to develop cameras which can meet the requirements of autonomous driving? We hope that we can go beyond the wisdom of animals in the natural world. And we know that we have many AI applications. Nevertheless, deep neural networks are hard to train and consume a lot of energy. For example, if we play chess, it could take a certain amount of energy. But let's look at an ant brain. It only consumes 0.2 milliwatts of power but can process many activities like nest-making, friends, fighting, or feeding and fights. So we're thinking, is it possible that we can develop our artificial intelligence that can help us to overcome the new challenges? And we also hope to transcend physical limits to enjoy truly immersive experiences. Currently, 5G networks are far from being able to offer such experiences, even though now the 5G networks have been deployed at a large scale. Therefore, we need to offer faster network speeds and lower latency for life-size holographic communication. We also hope we can extend our horizons by developing microscopic devices. What I mean by that is scientists can use computing to realize molecule and atom-level design and assembly so as to greatly improve the performance of chips and components. As we all know, our world is constructed on three pillars: matter, energy, and information. And from them, we can figure out what future challenges we will face. Matter is the origin of existence, energy drives motion, while information determines connections. We predict that by 2030, there will be hundreds of billions of connections worldwide. Broadband speeds of 10 Gbps will be available to every user. We will see a 100-fold increase in computing power and storage capacity. More than 50 percent of energy will come from renewable sources. Technologies around information and energy generation, transmission, processing, and consumption will need continuous evolution. Based on these predictions and assumptions, I would like to talk about the challenges we will need to address and the directions of the employment or further development in the next decade. The first challenge will be about defining 5.5G to support hundreds of billions of diverse connections. 5G is for enabling everything so that they can be connected, but in addition to connecting people, we also need to connect massive numbers of things, and the demands for those connections will be very diverse. The three use cases defined by 5G are not able to support all kinds of diverse IoT scenarios. For example, industrial IoT applications require both massive connections and large uplink bandwidth. So they need to develop a new scenario that falls between eMBB and mMTC, and we call it uplink-centric broadband communication, known as UCBC. There's also another type of applications that need ultra-broadband, low latency, and high reliability, and that means there will be a scenario that falls between eMBB and URLLC, and we'll call it real-time broadband communication, RTBC. Vehicle-road collaboration in connected vehicles requires both communication and sensing capabilities. So we need other new scenarios known as harmonized communication and sensing, HCS. Therefore, 5G scenarios must be upgraded from the previous triangle to the hexagon of 5.5G so as to support all things connected and all things intelligent. This is the second challenge. We know that the challenge posed by the connection via 5G is about the number of connectivities that we can achieve, but in terms of fiber capacity, it is very important. Today, a single fiber can support 1 million 4K video streams at the same time. We hope that 10 years from now, that is in 2030, it will need to be able to support 1 million people interacting in mixed reality, known as MR. This means that for the capacity of a single fiber to exceed 100 Tbps, and it will need to expand by 10-fold. First of all, we'll have to work on optical transceiver lasers and use high-modulation components to double or triple baud rates. In addition, new modulation, coding, and algorithms will be required to multiply the capacity. And second, we must develop new broad-frequency and low-noise optical amplifiers that support manual control for reliable auto long-haul transmissions. And third, we'll need to study dynamic controls for optical networks and transform the WDM networks into a synchronous system to improve anti-interference and efficiently use optical resources through computing. And of course, when we use those methodologies in the future, it will only be an increase of 10 times. And in the longer term, we also need to research new fiber and optical systems like space-division multiplexing. If you are in this field, then you will know that it's about applying the radio technologies into the fiber domain so as to increase the capacity of a single fiber by 100-fold. What about the third challenge? It's about network protocols. Today, our networks can support tens of billions of consumer connections. By 2030, they will need to support trillions of industry connections, and this will bring three major obstacles to network protocols. The first will be in achieving deterministic networks. As we know, for IP, the latency is undeterministic. Therefore, we need guaranteed deterministic latency. We must use new network calculus theories and protocols to transform the best-effort network latency we have today into a deterministic latency that can be calculated in advance. So that is one part of the challenge. And second is about security. When all things are connected, security systems will face serious challenges. Large numbers of external devices such as drones, cameras, edge computing devices, and sensors will create new security risks. They themselves are smart devices, and therefore, a large number of external devices will introduce new security risks. So we have to adopt intrinsic end-to-end security frameworks and protocols. Third is flexibility in the IP protocol. As the variety of industry requirements increases, some will require longer IP addresses while others will require shorter. To resolve this issue, we will need to expand IP addresses with fixed lengths and develop new internet protocols that feature semantic and syntax flexibility. So the third challenge is about phasing with the industry internet. We have to upgrade and optimize our network protocols. And now let's look at challenge number four: computing capacity. Capacity will need to determine the strength of the intelligent world, as we need advanced computing power to support the intelligent world. In 2030, the demand for computing power will see a 100-fold jump. The current single-core CPU performance increases about 50 percent every year, but right now that rate has dropped to 10 percent only. And we are also finding that general-purpose computing is very inefficient in some domains. Achieving sufficiently advanced computing power will be a huge challenge. First, digital computing will shift its base from general-purpose to special-purpose processing units. Heterogeneous computing architecture such as CPU, GPU, XPU will coexist in the base. Second, in the digital world that will live in, we might still want to leverage analog computing because analog computing will benefit certain fields. Optical computing will be applied to signal processing, combinatorial optimization, and machine learning. In particular, it will prove to be of great value for massive MIMO in wireless communication and optical communication. Therefore, optical, photonic computing, and analog computing are two interesting research topics. Challenge five: extracting knowledge from massive data for breakthroughs in industrial AI. To apply AI in vertical industries, diverse applications and trustworthiness of AI are inescapable problems. Someone asked the question, when will we realize the third generation of AI? It is indeed a challenge. AI models of common use are key to addressing these challenges. Using unlabeled data to expand the models from semi-supervised to self-supervised, the key is to build common AI systems. This area awaits breakthroughs. Second, the intersection of AI and scientific computing benefits common AI solutions. AI has provided new methods, tools, and thinking for scientific computing, while the latter contributes to the interpretability of artificial intelligence. Finally, trustworthy AI is a long-term goal. It is a great challenge for mission-critical areas such as those where lives are at stake, for example, autonomous driving, because not just correlation but causality must be made. Challenge six: storage. Storage density and ease of use are two challenges. First, storage density. With the same unit space and power, storage density will need to increase by 100 times in the coming decade. Confined by media and power, storage systems can accommodate such an increase. The future storage systems will need breakthroughs in large-capacity, low-latency memory tech such as DNA storage and high-dimensional optical storage, and new spatial model and coding tech to tear down the capacity wall. Second, ease of use. In the storage systems, data access speeds will increase from terabyte per second to petabyte per second, and access latency will decrease from milliseconds to microseconds, and its performance will need to increase by 100 times. Under the von Neumann architecture, data needs to move between the CPUs, memories, and media. However, the progress of PCIe and DDR bandwidth lags far behind that of the outside networks. The future storage systems need to be free from the von Neumann limitations. Instead of CPU-centric, they will be memory and data-centric. Instead of moving data around, they will be moving computing in order to tear down the performance wall. This is the sixth challenge. Now, challenge seven: combining computing and sensing for a hyper-reality multi-modal experience. The intelligent world should offer ultimate user experience. I think by 2030, hyper-reality will come true. So what is hyper-reality? It is created by seamlessly integrating the virtual and physical worlds. It first needs to accurately perceive and reconstruct the physical world, and then accurately understand the user intent in the combined worlds. To do this, the surrounding environment, we have to integrate information such as speech, touch, light, and brain-computer data collected by multi-modal sensors. The surrounding environment should act as a supercomputer that collects the information by relying on the above-mentioned sources of data. It analyzes user intent and presents information to users through naked-eye 3D displays, holographic projection, AR contact lenses, digital smell, and digital touch. Challenge eight: enabling continuous self-monitoring for more proactive health management. How to manage chronic diseases is a prominent issue for an aging society. Statistics show that 85 percent of deaths are caused by chronic diseases. That makes real-time monitoring essential. It requires breakthroughs in medical-grade wearables, such as those for non-invasive blood glucose and continuous blood pressure detection, as well as continuous ECG. Take blood pressure as an example. Optical sensors can provide more accurate pulses than PPGs and better data for blood pressure modeling and algorithms. Cloud services and artificial intelligence can be leveraged to build an integrated health data platform for individuals to implement proactive health management. Brain-computer interface, muscle-computer interface, and exoskeleton robotics and such can contribute to proactive health management and dignified, happy lives of the elderly. For one to be happy, one has to control and manage your own health through the BCIs, MCIs, as well as exoskeleton robotics, instead of handing over all your routines and tasks to the robots. Let me turn to one final challenge: energy. Targets of emission peak and carbon neutrality have prompted faster transformation toward renewable energy. Challenges of power generation, storage, and consumption have also been introduced. Regarding power generation, centralized power solutions are evolving towards distributed ones, which means the power system will be closer to the users. In the past, users were pure consumers, while they will be prosumers. In this way, there will be more energy nodes of bilateral transmission. Such a power grid shares the same characteristics with the communication networks. Regarding storage, the traditional grids barely store the energy. As renewable sources will play a primary role in the future mix, storage will be essential. That will add to the complexity of power grids. Affordable and smart energy management systems will be essential. In terms of power consumption, that will be key to building zero-carbon communities, campuses, and cities. A smart kinetic power grid is indispensable. It involves the following technologies for generation, storage, and consumption. Let me talk about the several key technologies. First of all, management technology. ICT technologies such as big data, AI, and cloud services integrated with the power grid to help manage watts with bits. Second, the control technology. The routers in the energy networks can be optimized to realize two-way energy flows and intelligent power distribution with a smart controller for the energy networks. Third, storage technology. New energy storage technology such as electrochemical and hydrogen will meet requirements of different use cases. Fourth, the basic technology of power electronics. Now, full-power semiconductor compounds including silicon carbide, diamond for medium and high voltage, and gallium nitride for medium and low voltages can help make energy devices more efficient and compact. These are the nine technical challenges and research areas for the entire ICT industry. We have addressed energy and health management, these two domains, what we engage in, and we have set goals for ourselves. We believe that energy will be one of the most considerable challenges for us, and it also represents great opportunities. What we hope to see in the intelligent world 2030 is stronger connectivity, faster computing, and greener power. We have to pull together our wisdom and innovation to address these challenges. Fusing humanity, Huawei is not working on this alone. We have to pull together innovation, and we have to do it in an open, inclusive, and collaborative way. We are building our labs, but at the same time, we are working with universities and research institutes in the joint labs. We are also providing funds for the research programs. These are just the various means of open collaboration in innovation. While at the same time, we are working with investors including the venture capital companies to provide funds for the cutting-edge technologies. In this way, we would like to incubate and develop new technologies. We embrace open, inclusive, and collaborative innovation mechanisms to overcome these challenges. With imagination, we can see the future, but with technology, we can get there. Industries, real-world problems, academic research, and venture capital, if well integrated, can facilitate collaboration and innovation that lead us into the intelligent world 2030. Thank you.

Thank you, Mr. Xu, for that insightful presentation. It is now my pleasure to welcome to the stage Mr. Eric Xu, Huawei Deputy Chairman and Rotating Chairman, to take us through Huawei's strategy during this era of uncertainty. Please welcome Mr. Xu to the stage.

Thank you, Mr. Xu. I would like to now ask you to stay on stage as we move to the Q&A component of today's event. It is now time for you in the audience to play your part. This is your chance to ask Eric a question. For those of you here in Shenzhen, please raise your hand and we will get a microphone to you. For those of you joining us online, please raise your hand virtually and unmute yourself when I call on you to ask your question. And those of you online, please feel free to also show us your face and turn your camera on. In both cases, please preface your question by stating your name and the organization you represent. We will try to get through as many questions as we can over the coming hour. I'll take the first question from the room. Madam in the second row, I believe you had your hand up first. Please go ahead.

Right now we've noticed many companies have announced their plans to make intelligent vehicles. Whether Huawei's strategy of not making Huawei-branded cars will remain unchanged? And if that's the case, whether it's a decision made in the context of the automobile industry being a long cycle, and Huawei made the decision in order to alleviate the pressure on cash flow and the chipset supply shortage? What are the specific information that you can share regarding Huawei's future plans for the intelligent vehicle business?

Okay, let's take the next question. There's a woman over there with her hand up. Can I ask you to stand up so the camera can get you as you ask your question, please? Please pass her a microphone so she can ask her question. Fantastic, thank you.

From China Business News. Last year you said the biggest objective for Huawei was to survive so that the company would continue to release its annual report for 2020. We've certainly seen the presentation of a Huawei annual report, and the business results were quite stable. You also mentioned it today that the external environment this year is going to be complex. So my question is, what are the challenges for Huawei this year, and what are your objectives for 2021, especially around areas like chipset and cloud services?

All right, I'd like to take a question now from the analyst side of the room. There's a gentleman there who I think had his hand up first. He's on the far right, about four rows back. And again, if you could please stand up so the cameras can pick you up.

There are two questions. First one relates to Huawei Cloud. We've seen frequent organizational restructuring of the cloud business at Huawei. It was previously a business group, and then it was changed to a cloud business unit where Eric yourself is the chairman. So where is the position of the cloud business in Huawei's portfolio, and what will be the strategic changes that we're going to see for this part of the business moving forward? Second question relates to Huawei's Kunpeng ecosystem. Just recently ARM announced its V9 architecture, and they said the architecture was not subject to the EAR of the US administration. So what is the progress in terms of licensing with Huawei? Whether Huawei has got a complete perpetual license for V9 architecture, and what are the plans and initiatives that Huawei will take to address chipset design and manufacturing challenges so that the Kunpeng ecosystem will continue to develop?

All right, I might turn now online to take some questions there. And as we do so, just a reminder to everybody who's joining us online to unmute yourself when I call on you, and to introduce yourself and your organization. Thank you to those who turn their cameras on, and if anybody else feels like it, please do so now. We have lots of people here on the call, and thank you for joining us. I will take my first question from Fanny Lee. Please, Fanny, please unmute yourself.

In recent years, we've seen many new intelligent vehicles with autonomous driving capabilities launched to the markets. We've also seen many high-tech companies jumping into this game. And with 5G, the vehicle has become part of the IoT interconnection and the application is booming. If we look at the car-to-car connectivity, car-to-road infrastructure connectivity, as well as the connection between vehicles and clouds, I have two questions for Eric. First, what are the changes 5G and the future 5.5G would bring to intelligent vehicles? Second, what are the recent developments of Huawei's Intelligent Automotive Solution business unit?

All right, I'd like to take another question online if we can. And again, thank you for turning on the cameras. Mark Sweeney, please unmute yourself and ask your question.

Thank you very much for giving me the opportunity to ask a question today. It's kind of two questions. The first one is obviously the US sanctions meant that Huawei has had to seek alternative supplies and has had to make its own way long before we've seen the pressure on the global supply of chips. I wonder, this problem is not going to go away for a long period of time. Demand is still surging. Is Huawei prepared? Will you be effectively immune to what we're seeing in the chip market because you have had to prepare in advance due to the sanctions? And my second question is, with Biden and the Democrat administration, I wonder might there be a change in relations between the US and Huawei?

All right, let me go to another question online. And just as I'm doing that, I'm going to remind everyone that we have fewer than 20 minutes left, so if you could please keep your questions brief. Okay, from Jakarta Globe, can you ask your question, please? No, the call was for Ibu Nada. If not, I think that's Mark Einstein from ITR. You need to unmute yourself, Mark.

The pandemic has accelerated the digital transformation in many countries. In our Asia Pacific region, countries are in the race for the leapfrog opportunity and to lead in the digital era. How do you see the future of digital transformation in the ASEAN-Pacific region? What role will Huawei play in this process?

All right, let's stay online and let's go back to Mark Einstein this time. Mark, are you still there? If you could unmute yourself, Mark. There you go.

Yes, thank you very much for your presentation and your update on Huawei's vision about 5G and 5.5G. However, my question is, given that we're talking about 2030, and many governments, operators, and vendors have already talked about launching 6G by 2030, I would like an update on Huawei's strategy and vision for 6G technology. Thank you.

All right, let's take one more question online if we can. Let me see who am I supposed to... ah, it's your turn now, Edem Gaddag-Beku. Can you unmute yourself, please, from African Fall? Ah, we may have lost Edem. Okay, I think we've lost Edem. I might go back to the room now. Can I get any... are there any takers for questions from the media to begin with? All right, the gentleman here in the front row. Can I ask you to stand up when you ask your question, please?

Let me confirm with you on the first question. It's basically how the cooperation between Huawei and the Japanese companies is going. So two questions: first, how Huawei plans to strengthen its cooperation with Japanese companies in light of the current situation? Second, Huawei's purchasing amounts in 2019 from Japanese companies was 1.1 trillion yen, and what is the number in 2020, and how that number would look like moving forward?

All right, we have time for one last question. I'm going to try and squeeze in one last question from an analyst. That gentleman there had his hand up very quickly. I'm going to ask you to stand up when you ask your question and to identify yourself and your organization, please.

From DBS Bank of Singapore. I want to ask a question in relation to HiSilicon. Now that HiSilicon has no access to advanced processing techniques, what will be the future profit-making model for HiSilicon?

And thank you, Mr. Xu. That was the last question of the day. Thank you very much. And this brings us to the end of our first day. Thank you all for your attention and for your questions. We will be back again for day two, starting at 9 AM China time. We look forward to seeing you again tomorrow.