Monday, 13 April 2020

What company is Tesla?

"Automobile companies" and "technology companies" will be the answer that many people will blurt out.  Tesla's automobile products are squandered around the world and are regarded as "catfish" in general, but few people will focus on Tesla's energy products.

 In fact, energy products have been written in Tesla's route planning.  In August 2006, Musk 's "Secret Car Plan" mentioned the Tesla route that is now widely known: it will make a small but expensive car to attract high-end users, and use the money to sell cars to develop lower prices.  Of medium-sized cars are sold, and capital is returned to make affordable mass cars.

 Musk also wrote eloquently that Tesla 's primary goal is to "accelerate the transition from a hydrocarbon economy to a solar and electricity economy."

 Unlike the route taken by electric cars, Tesla's solar road began with acquisitions.  "Tesla will sell other sustainable energy products together with other companies," Musk wrote in the plan, enough to prove that Tesla's vision for solar products is not 100% self-developed, and it is very likely that Tesla  It will also be a third party that undertakes brand sales.

This idea is being subverted.  After a lapse of 14 years, Tesla's detours on solar photovoltaics and the layout of energy storage products are far more than we thought.

 Save the Musk family wealth

 For solar photovoltaics, Musk seems too paranoid.  In 2016, Musk stood at the opposite of all shareholders and insisted on acquiring Solarcity, a home photovoltaic power generation company, for US $ 2.6 billion.

 At that time, Tesla's highlight moment, Model S / X has been verified by the market, Tesla announced the release of Model 3, the current star product, and ushered in the third quarter of the year.  The first quarterly profit.  The enthusiastic shareholders are not willing to take over the hot potato of Solarcity.

 The Musk called "a transaction without thinking", which caused Tesla's stock price to drop by 10% on the day of completion of the acquisition.

 What moved Musk was that from 2013 to 2015, taking advantage of the policy, Solarcity has grown into the largest solar photovoltaic roof company in the United States.  In order to reduce the consumption of fossil fuels, California launched the Self-Generation Incentive Program (SGIP) a long time ago to encourage more users to deploy power generation facilities by subsidizing electricity prices.  Under the government subsidy, the usage cost of Solarcity roof is 1/4 lower than the price of the public power grid, which has harvested the first wave of customers who received photovoltaics.

 The controversy of people standing on the opposite side of Musk is whether Solarcity's business model can continue: rent solar panels, implement the first free installation for users, and users pay solarcity long-term electricity charges.  The payment is very slow. In order to solve the problem of cash flow, Solarcity converted the lease contract into securities. However, due to the interest rate and interest rate increase, the company's bond value has been declining.

 At the same time, power generation subsidies are both temptation and poison.  Too much reliance on government subsidies to stimulate demand. In Solarcity's leasing model, in order to earn a difference in electricity bills, more installations are required, but it also amplifies upfront expenses.

 Before being acquired, Solarcity was actually insolvent. In 2016, the loss reached 820 million US dollars, carrying a cash flow of 2.6 billion US dollars, and the installation volume continued to decline.  Based on the expectations of shareholders.

 Musk's vision after the acquisition of solarciy is also a win-win situation: Tesla will transfer solar panels to store sales, reduce its own cost of sales, and stop the difficult leasing model; Solarcity can also help Tesla fill upstream  To solve the charging problem in the early stage of the popularization of electric vehicles.

 This "sorrow" is not understood by most people, and after so many years, the linkage of the two businesses is rarely shown to the public.

 More people will be willing to interpret it as a rescue operation for the Musk family.

 The clue is that at the end of 2015, Solarcity's market value had shrunk by about 60%, and Musk had held Solarcity's US $ 100 million bonds and 22 million shares as early as 2014, and became a major shareholder of the company.

 In addition, in addition to Musk's brother, Kimbal Musk, several members of Tesla's board of directors cross-share Solarcity and Space X. The founder of Solarcity is not an outsider. It is Musk's watch.  Brothers Lyndon Rive and Peter Rive-If Solarciy is allowed to go bankrupt, Musk and his family are undoubtedly the biggest "victims".

 Another focus around the outside and Tesla's board of directors is whether the valuation at the time of Solarcity's acquisition is too high.  This is one of the largest single acquisitions since Tesla was established so many years ago, which is equivalent to Musk buying most of his own assets at a high price with the hard work of everyone.  The lawsuit triggered by the financial rescue wrapped in the acquisition of sugar coatings has not ended yet.

 Another way to go, Musk will honor
 Solarcity's spring did not appear after being taken over by Tesla.

 Musk's idea of ​​integrating Solarcity business five years ago is not being fulfilled as expected-in Tesla's expanding charging pile system, in addition to a small part of the energy source is from photovoltaic power, more is directly powered by the urban power grid.

 After the acquisition of Solarcity, Tesla soon released a solar roof of its own. The photovoltaic power generation device is placed in the roof tiles. When the user lays the roof of the house, it is equivalent to laying a  Rows of photovoltaics.

 In 2018, Tesla also cut off several solar installation centers and withdrew its agreement to sell solar panels with Home Depot, the largest building materials manufacturer in the United States, which means that Tesla has actually indirectly abandoned Solarcity's once the industry's first position.  ——Cooperating with roof builders is the most direct way for every rooftop solar manufacturer to promote its business and sell products. The top three in the industry, "Sunrun" and "Vivint Solar", have adopted this cooperation.

 According to Renewables data, Tesla's residential solar installation market share has dropped from 33.5% during the peak period in 2015 to 9.1% in 2019.  Today, Solarcity still drags down Tesla's finances.



 The market share trend of Tesla solar roof, Sunrun and Vivint Solar, the top three PV companies in the United States

 Tesla is working hard to revitalize its solar roof business.

 After cutting off all third-party cooperation, Tesla was forced to "self-reliance" and vertically integrated solar roof sales in car-operated stores. Last year, it also resumed the original overthrowing solar rental service, and users only needed $ 50 per month.  You can install a solar panel system on the roof.

 Unlike before, with the sales of Tesla vehicles, Tesla's cash reserves have been driven, and financial parties have been introduced. The leasing model can also have solid cash support.

 Tesla did not hesitate to rare price reductions to recover photovoltaic customers, allowing customers to participate in product quotations, and finally finalized the price of 2 dollars / watt.  This price has fallen by 20% compared to the previous price, which is 40% lower than the national average price of residential solar systems ($ 2.85 / watt).

 A fire pressed the deceleration button of Tesla Solar roof in the United States.

 In June 2018, Walmart, a supermarket chain, publicly sued Tesla for choking to remove the solar roof installed in 240 stores because it found that a fire broke out even after the roof was disconnected from the system.  Wal-Mart listed several of Tesla's crimes: the electrical and solar systems were not properly connected, the product had obvious technical defects, there were no regular maintenance personnel, and lack of professional knowledge.

 Subsequently, Amazon, another important Tesla partner, also stated its position, saying that no more Tesla solar systems will be installed.

 All the safety issues are red lines. The voices of the two major business giants have pushed Tesla's solar roof business into an impasse. At present, many insiders predict that this business will be divested by Tesla again.

 At least Musk would n't think so. He slapped his head and announced on Twitter: Go to Europe and China.

 Tesla China disclosed in early March of this year that it was already recruiting for the development of solar rooftops in China.  A person familiar with the matter told 36 krypton, "solar roof has already contacted two listed companies in the photovoltaic industry in China and will purchase domestic components."

 However, this time when entering China, Tesla's photovoltaic business is far less fortunate than its electric vehicles in China. The domestic photovoltaic market is not completely blank. The penetration rate in third- and fourth-tier cities is very high, and the cost has been able to be extremely low.  .

 According to his prediction, Tesla may next mainly target high-end users with villas and roofs, supplemented by small industrial and commercial enterprises, and rural users are not Tesla's target users.

 For wealthy high-end users, energy conversion rate and installation time are the key factors that affect the choice.

 "Recently the domestic photovoltaic market has fully recovered, and Tesla's pricing will definitely be higher than domestic competitors." He has reservations about whether Tesla's photovoltaic business can be well implemented in China. He believes that "special  Sela 's marketing and brand are very powerful. It may be bundled with electric vehicles. The profit of photovoltaics is not high. Tesla 's photovoltaic brand premium must reach at least 10%, otherwise it is difficult to make profits.  Hard to say."

 Another bargaining chip: Tesla Power Bank
 The prospect of solar photovoltaic roofs is not yet clear, and Tesla has actually dropped a few early on the energy board.

 What is the biggest problem of solar energy?  Musk's answer is-"The sun does not come out at night."  (The obvious problem with solar power is the sun does n't shine at night.)

 In April 2015, Tesla has released a series of plans for solar energy storage products, including Powerwall's household battery energy products, which can be charged when the demand for electricity is low, and when the price and demand for electricity are high  Output electric energy, the maximum stored energy is 10 kilowatts, which is equivalent to the electricity of an ordinary family working continuously for 10 hours.

 In addition to home use, Musk 's ambition is to expand the PowerPack and advanced Megapack for commercial and public energy storage products. The idea is very simple. Multiple Powerwall components are connected in series, and the basic power can reach 100  Times (10,000 kW) or even higher.

 Musk once boasted that "only 160 million Powerpacks can cover electricity in the entire United States" and "the super energy storage equipment formed by multiple Megapacks in series is enough to provide 6 hours of electricity for every family in San Francisco."

 Solar capacity + energy storage + electric vehicles as power terminals, Tesla's layout is also an energy closed loop.  But at present, it seems that the priority of energy storage products will be higher than solar roof.

 Compared with solar roof's price cuts, the price of entry-level Powerwall has been rising.  Although it has risen from $ 5,900 to $ 6,700 in recent years, it cannot stop Tesla from setting a new record of 415Mwh energy storage deployed in Q2 in 2019.

 Moreover, energy storage can resonate better with Tesla's car making than capacity: Powerwall also uses a thermal control system, battery pack and architecture similar to the Tesla Model S.  In addition, with the increase of electric vehicles and the growth of service life, the recycling of electric car batteries in the future is also a problem that many car companies are now tackling.  The idea of ​​using the batteries of scrapped vehicles in Tesla's energy storage products is very reasonable. Nissan and BMW have done this before.

 Of course, with the habit of making the best use of Musk, he is also increasing the coordination of the automotive business and the photovoltaic business. In addition to installing solar roofs in several super factories, Tesla 's first electric pickup, Cyberrtuck, was launched last year.  For the first time, the car is equipped with solar panels.

 Although car companies do not start with solar panels in cars, Toyota Prius and Hyundai Sotana Hybrid are equipped with solar panels that charge the hybrid system and the top cover. These solar panels are at best icing on the cake.  Solve the energy problems of vehicle components.

 Tesla previously said that the optional Solar Panel has increased the endurance of Cyberruck, which will also make it the first mass-produced solar car.  According to official data, as of April, orders for this car have exceeded 600,000.

 Whether it is an SUV, a sedan, or a smaller and more subdivided pickup truck, Tesla's electric vehicle growth trend is no doubt, so what is the next potential growth point?

 Perhaps it is energy.  The silence of the past decade is due to the difficult climb in Tesla 's production capacity. All forces are concession to the main channel of the automotive business. By the beginning of 2018-2019, in order to support the capacity growth of Model 3, Tesla still needs  Reduced energy storage package products and roof production.  However, after a short time, Musk jumped out and shouted, "2019 is the year of solar roof and Powerwall."

 Since last year, there have been more and more actions on these two product routes: in 2019, we cooperated with Pacific Gas and Electric Company (PG & E), one of the largest power companies in the United States, to deploy several Megapacks in California; an advanced version was launched  Solar roof V3 has improved greatly in terms of installation efficiency, cost and capacity efficiency. In March 2020, Tesla said that it will build 224 Megapacks on the island of Hawaii. After landing in 2022, it will be the latest giant in the world.  Battery system ...

 Tesla's blueprint for those rich fans is like this.  Install a Solarcity solar roof that can convert solar energy into electrical energy on the villa, store the electrical energy in the home battery Powerwall, and charge your own Tesla electric car.  Is the car running out of electricity?  Not afraid, the solar panel on the car can help the car to recharge at any time.  Power failure during peak hours?  Not afraid, the owners of their own "microgrid" for electric cars and batteries can also sell electricity back to the public grid.

 From this perspective, in the future, Tesla is not an automobile company, but an energy company.

WILL PICO SURVIVE?

In 2020, the rapid development of 5G has brought hope for profitable VR headset startups who have long burned money. However, the sudden outbreak of the Coronavirus epidemic accidentally poured a basin of cold water.

In an exclusive interview with 36kr, Pico founder Zhou Hongwei said that Pico originally planned to achieve profitability this year, but due to the epidemic, there is great uncertainty as to whether this goal can be achieved. The epidemic affected not only Pico's domestic market, but also overseas markets. If the June epidemic can be controlled globally, Pico will still have a chance in the second half of the year.

However, Zhou Hongwei believes that Pico has now stood at a better point in time. Even if it is not profitable this year, it will only lose a small part, and it will definitely be profitable next year.

Recently, Pico Neo 2 officially launched for sale. This is a VR headset that supports 6DoF tracking of the head, and benchmarks Oculus Quest. 6DoF refers to 6 degrees of freedom. Compared with 3DoF, there are more degrees of freedom in up, down, forward, and left and right. Therefore, in addition to turning the head in different directions, it can also detect the spatial displacement of the head. This way you can move freely in the virtual reality world. 3DoF VR devices are more suitable for watching videos, and 6DoF devices are more suitable for playing VR games.

The release of Oculus Quest last year brought the 6DoF VR all-in-one to the fast lane, driving the development of the VR game console ecosystem.

Zhou Hongwei believes that the Oculus Quest application experience will become a mainstream direction for VR users. Oculus sales will grow explosively, and 6 DoF VR game consoles will also have a rapid growth.

In the next three to five years, the domestic VR game console market will grow exponentially, from tens of thousands to hundreds of thousands, or even millions of units a year. The market volume climbing to 500,000 to 1 million units will be an inflectionpoint. "In fact, from 1 million to 3 million and 5 million is very fast, but climbing to 500,000 to 1 million in front will be a challenge. Depending on the product experience, brand, marketing, and content, traditional Internet or The involvement of some giants of the gaming platform, either investment or cooperation. "

But for now, the high price of more than 4,000 yuan is not close to the people, and the choice of first-generation content is relatively limited. Zhou Hongwei expects Noe 2 to have sales of 50,000 units this year.

Although early sales expectations were not high, Zhou Hongwei believes that this will not affect the enthusiasm of overseas developers for transplanting this headset into Chinese. Many high-quality VR games have already made money in Quest, and the cost of porting across platforms is not high. It only takes one or two engineers to work in three or five days to complete.

Most importantly, China is a very promising market. Quest users are growing at a high speed. SuperData data shows that Quest shipments reached 705,000 units in 2019. Zhou Hongwei predicts that Quest users may reach 10 million in three years. Correspondingto the Chinese market, it should be about 1/2 or 1/3 in the future. China will also have an ecosystem of millions of users, which is still attractive to developers.

The following is a conversation with Pico founder Zhou Hongwei in an interview with 36 kr, edited by 36kr:


New product
36kr: Recently we released the new product Neo 2. What is your expectation on the sales volume of this product, and how much will it reach in 2020?

Zhou Hongwei: I am very optimistic about the future big market, but for this year, I dare not expect the domestic market to be too high. The price of more than 4,000 is not an entry-level product, and it is a small circle in the entire user base. Players who are willing to spend 10,000 yuan or 20,000 yuan to buy a computer or game book, these people may spend more than 4,000 yuan to buy a VR is not a problem, but this circle will be relatively small. I personally think that there may be 50,000 units this year, and more than 50,000 users are already good.

36kr: In fact, the sales of 50,000 units is still a very niche market.

Zhou Hongwei: It is the first-generation product. In addition, the content of the first-generation product is relatively limited. There is no particularly good content for you to choose. From the perspective of this industry, Oculus Quest is the first year in 2019, and Pico Noe 2 is the first year in China in 2020. We think that such a high price in the early stage compared to the first-generation product has this amount (still okay) . If the price of this product quickly drops to around 3,000 yuan, and the content is extremely rich, I think it will soon be able to amount to 200,000 to 500,000 units.

36kr: If the price reaches 3000 yuan, can we still make money?

Zhou Hongwei: If it is definitely not profitable to 3000 yuan now, but in the future, I think with the increase in volume and cost optimization, even if we sell more than 3000 yuan, we still have a certain profit, which is possible.

36kr: Domestic 6 DoF headsets with annual sales of 1 million or 10 million units, what kind of conditions need to be met?

Zhou Hongwei: I personally think that the market for VR game consoles in the next 3 to 5 years will be an exponential curve of rapid growth, from tens of thousands of units to hundreds of thousands of units in a year to millions and millions of units. When is a turning point in quantity? I think it might be 500,000 to 1 million. In fact, from 1 million to 3 million and 5 million is very fast, but climbing to 500,000 to 1 million in front will be a challenge. Depending on the product experience, brand, marketing, and content, traditional Internet or games may be required. The involvement of some of the platform's giants, either investment or cooperation, we will look at this path, but we are relatively certain that this market demand is there.

For example, I predict that the Oculus Quest platform will reach 10 million users in three years. Corresponding to China, 10 million overseas, China should reach 2 to 3 million. There won't be any special ones in this market, maybe just one or two, and we certainly hope that Pico will be in it. At that time, did Pico introduce other large Internet companies from the perspective of capital or cooperation, and we are also working hard or thinking in this direction.

How can I get to 500,000? The first is to do a good job of the product and experience, so that the future can only be achieved by users. The second is to control the cost well. For example, in the future, if the cost is about 3000 yuan or about 3000 yuan, you will have the opportunity to reach 500,000 units. In addition, there must be a lot of good content, which is supported by word of mouth and marketing.

36kr: Do we now seek such cooperation with giants?

Zhou Hongwei: There are a lot of exchanges, but I think the focus of their entire strategy is not necessarily prioritizing VR so high. But as long as the overall entertainment trend (for the better), especially if overseas Quest can sell 5 million units, including the content of the game, forming a positive earning ecology. Some domestic game and Internet giants, I think they will definitely pay attention and invest. We hope that at that time Pico will be the best, whether it is a collaborator with them or an investment target.

Epidemic
36Kr: Under the influence of the epidemic, have the companies resumed work now?

Zhou Hongwei: Some people are working from home, some are in the office, not everyone can go to the office, and the Qingdao office has all resumed work. The return of the Chinese New Year in February was affected by some shortages. Some of the next batch of goods supplemented production in February. At present, China's impact is relatively small. Domestic OEMs, including our foundry companies, have basically resumed work.

36kr: How will the epidemic affect the promotion and sales of new products?

Zhou Hongwei: In terms of promotion, one is that it affects us offline, and the offline ones have basically stopped. Another one has a great impact on the pace of our entire product release. Before the Spring Festival holiday, we also planned to say that there will be a press conference in February, but we found that the situation was very bad at the end of January. The next press conference is also changed to this online media promotion, including some live broadcasts in e-commerce, similar to Tmall and JD.com, and of course, like games. In general, online promotion has been strengthened and offline promotion has been weakened.

From a sales perspective, offline must have been greatly affected. In addition, a large part of domestic To B, especially education-related, has been significantly negatively affected in February and March because schools are not open. Fortunately, most of Pico's sales are done through online To C. Online sales have never stopped. Online sales are better than when there is no epidemic. The data of user activity and duration of use are actually Better.

36kr: Has the epidemic brought some new demands and opportunities to the VR industry?

Zhou Hongwei: We are seeing some new requirements for VR's new gameplay. For example, can we take a class in a virtual classroom, with students at home and teachers at home? With VR, a virtual classroom can be realized. In the future, our product line in To B is also developing some new versions internally to support everyone to use VR to use virtual classrooms. This was originally a direction our To B was doing, but now it is speeding up.

In the next one to two years, I think that due to the epidemic situation, this kind of virtual and more immersive VR application on the line should be promoted, which will lead to many opportunities.


36kr: Has Pico made money since the company was founded? How much has the epidemic affected progress towards profitability?

Zhou Hongwei: We haven't made any money so far. We originally expected to make money this year, but this year's epidemic is really annoying. It not only affects the domestic market, but also greatly affects our overseas market, including the market of B and the market of operators. This year is generally not optimistic. We originally planned for this year at the end of last year, or we want to make a profit this year. Now we can say that there is a lot of uncertainty. If the entire epidemic is controlled worldwide in June, and then in the second half of the year, if you drive the economy or high investment, Pico will still have a chance. I said that there is still a chance.

36kr: What is the proportion of domestic and foreign markets for Pico VR headset shipments?

Zhou Hongwei: In terms of our G2 last year, domestic and foreign are almost the same, because overseas operators bought a lot of G2 things. From this year, there should be more domestic.

36kr: If the situation changes, do we have Plan B?

Zhou Hongwei: We have adjusted, we are still controlling costs, and we are also controlling costs for profit this year. At present, our adjustment is also fine-tuning. Even if we have not made a profit this year, the result may be a small loss of money, or it is basically close to supporting ourselves. In fact, in this industry, it is already very good in this special year. We don't want to lose such an opportunity. We think that even if it is affected this year, VR will grow at a high speed next year. I think we will definitely be profitable next year.

In such difficult times, the more difficult it is, the more we must adhere to the quality of our products and be sustainable. We do not want to lose such opportunities because of the epidemic, because we reduce our investment. We are now standing at a good point in time, just need to go to the market and give us a chance, give me a boost, we are a good company with positive cash flow in this industry.

VR content, developer ecology
36kr: How to solve the problem of too little VR content?

Zhou Hongwei: Many games with good experience have slowly come to VR, and many are already on the way. In the past year, many studios, including some very small ones, that have switched from traditional games to VR games, have made money quickly by developing VR games, and they have made good money, and they are not one. For example, Valve Software (the developer of the VR half-life game) has made tens of millions of dollars, and small ones have made hundreds of thousands of dollars and millions of dollars. We have already seen positive content for VR. A boost, many VR content developers are willing to invest in VR games, because he may foresee a good return.

36kr: For overseas developers, if our Neo 2 sales are about 50,000 units, plus the conversion rate, can this offset the cost of Chinese implantation?

Zhou Hongwei: In fact, all VR is better, the logic behind them is that they are optimistic about the future of VR games. These good games should have made money in the Quest platform. They actually have to make a choice here. I think the logic behind them is to be optimistic about Pico's entire product, not just from the perspective of 2020. The content should consider the next 2 to 3 years. There are users in China, and the growth of Quest users from tens of thousands to 700,000 to millions, if it falls to China, it should be about 1/2 or 1/3 in the future. China will also have an ecosystem of millions of users, which is still attractive to him. As far as this market is concerned, its investment in transplantation is still small.

In addition, the cross-platform migration of such content is actually a relatively small workload, basically one or two engineers work for three to five days. Of course, we have provided good support to reduce the cost of his transplant and improve the overall efficiency. Combining these two aspects, a developer who spends three or five days working as an engineer can be ported to a potential platform in China. I think it is still attractive.

36kr: What kind of payment model do we have with these overseas developers?

Zhou Hongwei: Generally speaking, it ultimately depends on sales.

What 5G means to VR

36kr: 5G's experience with this VR headset and what changes can the entire VR industry bring?

Zhou Hongwei: There are two kinds of relationships, one is the significance of 5G to VR, and the other is the significance of VR to 5G. In the past year or so, when global operators launched 5G, they were looking for things that could drive 5G trafficconsumption. VR is currently the most direct and fastest technology for 5G to land quickly and expand users. Whether it is from several major operators in South Korea to the three major operators in China, they have clearly proposed to bring VR into 5G applicationscenarios to expand more 5G users and also expand VR users.

On the other hand, will 5G help VR? In fact, the current S-class products are mainly WiFi, and 5G is not used directly. In the future, we think that VR with 5G connectivity is an inevitable trend, but it is not currently.

5G or Gigabit broadband still improves the VR experience, mainly in bandwidth. I am optimistic about a trend now. In fact, 5G mobile phones are used as a standard terminal and VR accessories are used. We think this is more related to 5G. We also seethat operators are fully pushing the integration of VR products in the form of Glass glasses with 5G mobile phones. We believe that the popularity of 5G users and the growth of 5G mobile phones will promote the rapid uptake of VR products in this form. From both perspectives, I think it's a mutually reinforcing situation.


36kr: The VR Glass released by Huawei is in line with the trend of VR accessories, but the response after this product was released was not very good. What do you think is the problem?

Zhou Hongwei: I think there are three aspects to the problem. One is that the experience itself is still lacking. After all, it is the first generation product and can have an understanding and tolerance. The second price is too expensive, such a price is not acceptable to the mass consumers. The third is the content of VR. Huawei is a very strong hardware or technology company, but the content of Huawei's entire VR is still relatively limited. Several factors have caused the first-generation products to sell poorly. But we should not deny this trend in the future.

36kr: How do you think Huawei, a mobile phone giant, has entered the VR market? How to meet their competition?

Zhou Hongwei: I think Huawei's VR has its inherent logic: Huawei itself is a communications company, a 5G company, and Huawei's first demand for VR is that it has so many 5G operator customers around the world, operators have a problem 5G base stations have been built, how can 5G bring benefits to operators?

Huawei has been recommending VR. In the early days of 5G, VR is a path that can help operators attract users, drive traffic, and increase sales. Therefore, it is not surprising that Huawei does VR. In terms of competition, I think Huawei's focus is still on the system and mobile phones. VR will not be a big strategy. This requires a dedicated team to do everything from technology, products, and content. And we see Huawei The product form is connected to the mobile phone in this way. I don't think Huawei will make products such as all-in-ones.

36kr: Is there a possibility of explosive growth of VR in the C-end market?

Zhou Hongwei: We have seen an explosive growth trend, driven by two factors, one is 5G, or a channel such as 5G operators is driving the growth of VR users. Another driving force is VR game consoles. In fact, behind a large game console ecosystem, the biggest driving force is Oculus. Everyone will soon discover that Oculus Quest application experience will become a mainstream direction for VR users, Oculus sales will grow explosively, and 6 DoF VR game consoles will also have a rapid growth.

Thursday, 9 April 2020

SpaceX's Rival OneWeb Bankruptcy: Outbreak Caused Directly; Many Top U.S. Universities Form Alliance with Microsoft: Supercomputer and AI Overcome New Crown

[Titanium Media Comprehensive] News on March 29. Recently, OneWeb, a satellite broadband company invested by Japan's SoftBank Group, reported bankruptcy application. According to the latest news from foreign media, the company explained in detail the reasons for applying for bankruptcy protection in an official statement.

OneWeb was founded in 2012 by technology entrepreneur Gregory Thane Wyler. OneWeb wants to provide people around the world with affordable high-speed Internet access by deploying hundreds of low-orbit satellites, which are also needed for data growth from 5G, the Internet of Things, autonomous driving, and other new technologies.

Foreign media said that the company's bankruptcy was unexpected because they were competing with Musk's SpaceX company to provide online services for people in remote areas. A week ago, the company sent nearly 34 satellites into orbit, bringing the total number of satellites to 74.

However, launching satellites is a costly endeavor, and OneWeb's own hematopoietic capacity is insufficient, relying heavily on external financing. OneWeb has raised funds many times since its establishment: In 2015, it received US $ 500 million in financing from investors such as Qualcomm, Virgin Group, and Airbus; in 2016, OneWeb raised US $ 1.2 billion, of which SoftBank Group invested US $ 1 billion. But at present, SoftBank is unable to continue to invest in OneWeb due to various debts, and the company is still facing a huge funding gap.

A few days ago, OneWeb CEO Adrian Stocker took the ongoing New Crown Virus pneumonia as a "Chapter 11 bankruptcy application" (not a liquidation company, but to provide bankruptcy debt protection) to the US Bankruptcy Court in the Southern District ofNew York. reason.

OneWeb introduced that the company has been seeking new funding since the beginning of 2020, but has failed to finalize the transaction. "Although the company is close to obtaining financing, the process has not progressed due to the financial impactand market turmoil associated with the spread of the new crown virus . "

Titanium media sound:

We know from studies of other coronaviruses that viruses can acquire this mutation and then they become more pathogenic. This explains why the virus spread so easily and caused the current pandemic. Surface protein mutations may have triggered the pandemic,but the weaker version of the virus has been circulating in the population for years, even decades, before accumulating to the present.

Our analysis, and others, point to an earlier origin than that (Wuhan). There must be some cases in Wuhan, but by no means the source of the virus.

-Robert Gary, Professor, Tulane University School of Medicine

It is extremely painful to implement high social isolation. I call it "closed." There is no middle choice. It is difficult to say that you will not infect others or be infected. Until we determine that the number of people infected is at a minimum, even if you tell everyone you can buy a house, a car, a restaurant, etc., I still very much doubt that people will really do it. People want to protect the elderly and their parents.

In short, the sooner we "take medicine", the sooner the highest-level medicine, the sooner we will get out of trouble and never return to it again.

--Bill Gates, founder of Microsoft


Tuesday, 7 April 2020

ALI's next move: online education APP, Banban Answer

Editor's note: This article is from the micro-channel public number "Tech Planet" (ID: tech618), Author: Chen Qiaohui

Under the epidemic, online education has always been a hot industry today. Answering and counseling products such as homework help, little ape search questions, and Xue Bajun are all popular, and major Internet companies are stepping up their efforts in this area.

Tech Planet (WeChat ID: tech618) was exclusively informed that Ali recently launched a product called "Help Answer", a paid question and answer platform for primary and secondary school students' academic problems. This is the first independent product launched by Ali in addition to the online education features launched by Ali.

According to the sky eye inspection, the developer of "Bang Bang Da" is Zhuoyi Changyou (Beijing) Technology Co., Ltd., and the suspected actual controlling shareholder is Alibaba Culture and Entertainment Co., Ltd.

Tech Planet (WeChat ID: tech618) found that "Help and Answer" can get course tuition through online payment. Supports quick explanations on topics that you do not understand by taking pictures, voices, videos, etc. If you don't have a problem, you canpublish a reward through the app to let more professional teachers teach online.

Open "Help and Answer", and you can find that the login methods are mobile phone number, Alipay, and WeChat. Remove the QQ, a third-party login method that is more popular with students, which is quite surprising.

After registration, there will be two options, one is the questioner and the other is the answerer. The user selects the identity and fills in the basic information to officially enter the app. At present, the scope of the questioner's identity is limited to primary school to high school students, and the scope of the answerer's identity ranges from undergraduate to postdoctoral, which means that the answerer is not limited to whether or not he is a teacher, but to answer the question with the undergraduate as the starting point.

Taking the questioner as an example, you can choose to take a picture or write a picture with a narrative to write down the question, and then select the grade corresponding to the question, write a commission, and post it. Questions posted need to be reviewed. If no one takes orders within 10 minutes after passing the review, you can get answers for free.

As the answerer, the homepage will display the questions to be answered, for the answerer to choose. After entering the answer interface, the answerer can discuss in words or record a video to answer. It is worth noting that the knowledge points involved in the related questions need to be evaluated before solving the questions, and the answers can be obtained after the assessment is completed.

The questioner checks the answer. After passing, the system will give the responder a certain percentage of commission and evaluation. In order to regulate the answering attitude of the respondents, the platform has established a credit score system for the respondents, and the quality of the answers is linked to it. If the credit score is 0, it cannot be restored and the account will be blocked.

At present, the "help and answer" is still in the testing stage, so the entire product is very simple in terms of design and function. Compared with the online education brand "job help", which is also a competitive product, there is less parent-level supervision, live answering and Community functions, it is not ruled out that such functional requirements will be increased in the later stage.

According to AiMedia Consulting's analysis of domestic online education users, the number of online education users in China will reach 305 million in 2020, and the new education model of "Internet + Education" is continuously expanding. As various onlineeducation platforms continue to develop and sink the market, the basic market of the online education market is directly expanded, and the number of online education users will further expand in the future.

Under the guise, Ali's introduction of "help and answer" also took advantage of the trend. This time, online education is in the second spring of the industry, and changes in the industry pattern after the epidemic are inevitable. In the end, who will come out on top still needs users to choose.

Wednesday, 12 February 2020

WHERE DO ALPHABET GO

Among Alphabet's innovative R & D subsidiaries, as some business failures are closed, some promising companies introduce external companies for independent development, and some companies that have a close relationship with Google's internal business are absorbed and reorganized. Is it necessary for Google to establish  An independent "Alphabet holding company" to package these R & D operations may require a question mark.

 A few years ago, the United States Google Corporation restructured and re-established a holding company called "Alphabet". In addition to Google, Alphabet mainly includes some innovative research and development companies that develop future businesses.  The emergence of Alphabet has also caused some confusion among the general public about Google's perception, and the media have to distinguish between Alphabet and Google.

 According to the latest foreign media news, with Pichai as CEO of Alphabet and Google, there have been some obvious changes in the company.  Recently, another Alphabet subsidiary was absorbed into Google. An American authoritative media said that these actions made the outside world question the future of Alphabet (if necessary).

 Why the architecture is reorganizing
 Why did Google suddenly become Alphabet?  This comes mainly from Google's innovative R & D business.

 As we all know, Google is a company with a strong technology and engineer culture. In addition to traditional businesses, the two founders of Page and Brin focus on various future projects, such as autonomous driving, life sciences, drones, satellite communications,  Robots and more.  Google has entered these innovative R & D fields through a large number of mergers and acquisitions and talent recruitment. These departments are also known as Google's "bet business."

 However, Google's continued investment and opacity in the area of ​​innovation and research has raised concerns among Wall Street analysts and affected Google's stock price.  Investors also don't know which Google's innovative research and development projects have future prospects, and which projects consume a lot of funds.

 To reassure Wall Street analysts.  Google restructured the company structure, newly established a holding company Alphabet, and arranged all its innovative R & D business in Alphabet. At the same time, Google continued to exist, retaining the company's pillar businesses that can earn real money, such as  Web searches, online advertising, YouTube video sites, Android operating systems, and more.

 However, Alphabet's innovative business is still difficult to get rid of outside doubts, many projects are progressing very slowly, and no hope of success can be seen.  Investors believe that Alphabet should not continue to throw money at projects without hope.

 Absorb again

 According to foreign media reports, recently, Alphabet's subsidiary Jigsaw has been merged into Google.  The company is committed to solving global security challenges through technology.

 The company was founded in 2010 by Eric Schmidt, a former Google CEO (later chairman), who used the name "Google Creative".  In 2016, Jigsaw left Google and became a subsidiary of Alphabet.

 The company's existing products mainly address issues such as online information review, crackdown on fake news, crackdown on online sexual harassment, and extremism.

 The company said in a self-introduction: "Our world has so much Internet content, and the risks of digital security have never been higher. We are working hard to find new problems that threaten our society-from the review and harassment of online information to false  Information and violent extremism. Through field research and cooperation, we strive to find new solutions to make everyone safer. "

 The company's best-known free service is the Shield Project, which protects news publications, civil society, election surveillance, and political campaign sites from distributed denial-of-service attacks by cyber hackers by leveraging Google's global infrastructure.

 The company also provides tools for people to identify fake pictures.

 According to reports, Jigsaw has quietly moved from Alphabet to Google in the past month, and it still operates as an independent unit, with daily operations unaffected.

 Jigsaw CEO Jared Cohen-formerly in the US State Department's policy work-now reports to Kent Walker, senior vice president of global affairs.  In addition, it may be the smallest Google subsidiary with about 60 employees.

 Previous merger

 Jigsaw is not the first time Google has absorbed a subsidiary of Alphabet.

 In early 2018, Nest, an independent company specializing in the development of smart home products (acquired from the outside for $ 3.2 billion, which is the second largest acquisition in Google's history) was merged into Google and joined Google's hardware division.

 It is well known that after the sale of Motorola Mobile to Lenovo Group, Google has developed a strong interest in the hardware business.  Google spent $ 1 billion to acquire the mobile phone foundry business of Chinese smartphone maker HTC, and gained 2,000 outstanding engineers.

 In the past few years, Google's hardware products have been increasing, such as smartphones, smart speakers, laptops, set-top boxes (including video sticks) and other products. These hardwares have also become some of Google 's Internet services (especially Google Assistant).  Landing channels have promoted Google's core network services.

 In addition, in 2019, Alphabet's cyber security subsidiary Chronicle also joined Google's cloud computing business.

 In the emerging cloud computing market, Google is far behind leading companies such as Amazon and Microsoft, and its market share is only a fraction of them, which makes Google quite awkward.  According to foreign media reports, Google has set ambitious goals to enter the world's top two cloud computing markets within a few years, that is, it must exceed Microsoft.

 Recently, Google is also enhancing the product line and competitiveness of its cloud computing business through various acquisition transactions.

 New officer

 At the end of last year, Alphabet made major personnel adjustments. The two founders, Page and Brin, who have long been separated from the actual management of the company, resigned from the positions of CEO and president of Alphabet respectively. In the future, Google CEO Pichai  He is also the CEO of Alphabet.

 Since Pichai led Alphabet and Google, there have been some significant changes.

 For example, recently, Alphabet announced the fourth quarter of last year's financial report. As the first time in history, the company announced detailed operating data of two important sub-businesses of Google-YouTube online video and cloud computing, which has been widely accepted by Wall Street analysts.  Welcome, and consider an important step in management transparency.

 Of course, some analysts are not satisfied. Some indicators of YouTube and cloud computing business lag behind competitors. The analyst hopes that Pichai will release key indicators such as profit margin in the future.

 In addition, Google 's image has plummeted in the past year or two. In addition to following Facebook 's countless scandals that violated personal privacy, various negative news also appeared within the company.  Employees are dissatisfied with management and some policies (such as Rubin, the father of the Android system who received the scandal, received a $ 90 million termination compensation). There have been many strikes and the company is losing the trust of employees.

 Recently, Google has also begun investing more resources in investigating high-ranking sex scandals.  In addition, according to the latest news from foreign media, Google 's human resources leader has been confirmed to be removed from the job because employees are strongly dissatisfied with the human resources department 's inaction on sexual harassment complaints.

 Is Alphabet cancelled?

 Alphabet's innovative R & D department has become a burden for the company.  In the past, Alphabet has closed down several failed R & D operations, such as satellite Internet plans and robotics (Alphabet has sold several acquired robotics companies, including Boston Dynamics, a well-known online company).

 Recently, Pichai said in an interview with foreign media that for these innovative R & D businesses, Google will take a similar approach to the life sciences business, that is, let them further operate independently, at the same time let external companies take shares, and at the same time take  Stricter audits.

 After Pichai became the head of Alphabet, industry insiders analyzed that Pichai could gradually absorb Alphabet's subsidiaries into Google.

 It is obvious that among Alphabet's innovative R & D subsidiaries, as some business failures are closed, some promising companies introduce external companies to develop independently, and some companies that have a close relationship with Google's internal business are absorbed and reorganized.  It is also necessary to set up an independent "Alphabet Holding Company" to package these R & D operations, which may need to be marked with a question mark.

 In Silicon Valley, a holding company structure like Google is still relatively rare.  Technology companies are more focused on flat management, and the company communicates quickly from top to bottom, and iterates quickly based on market demand. More management levels mean lower communication efficiency and administrative costs.

Monday, 3 February 2020

350 million acquisitions? Nervana gets hacked, Intel focuses on Habana

After acquiring two cloud AI chip startups in succession to consolidate its most important data center market, Intel finally made a choice between the two tigers.

Habana beat Nervana for its excellent scalability technology and design, and its ability to deliver in volume.  But this does not mean victory in the end.  For Intel, as long as he sees the opportunity to surpass Nvidia, he will not hesitate, the key is only to choose the right chip.

Recently, foreign media broke the news of Intel's new actions in the field of AI.  Forbes said that Intel will stop TSMC 's 16nm foundry Nervana chips, mainly the Nervana NNP-T series.  Intel did not announce the matter with much fanfare, but has informed analysts and customers via email:

The development of the Nervana NNP-T training chip (formerly codenamed Spring Crest) has been stopped, but the promise made to customers by the inference chip Nervana NNP-I (codenamed Spring Hill) will be honored.

The Habana product line offers the powerful strategic advantages of a unified, highly programmable architecture for both inference and training.  By moving to a single hardware architecture and software stack to enable data center AI acceleration, our engineering team can work together to deliver more innovation to our customers more quickly.

On Friday, deep learning analyst Karl Freund also tweeted that Intel will completely stop the NNP-T products of deep learning chip startup Nervana that Intel acquired in 2016 and focus on Habana Labs.

1. Why buy two cloud AI companies?

 Although Intel has layouts in the AI ​​field for terminals, edge computing, and the cloud, the top priority is still to take advantage of Intel's strong position in cloud data centers to ensure its position in the cloud AI chip market.  This is also the important reason why it successively acquired two cloud AI chip companies, Nervana and Habana.

 On December 16, 2019, rumors of Intel's acquisition of Habana Labs were finalized. This acquisition continues Intel's "big-handed" style-$ 2 billion, second only to Mobileye's second largest acquisition.

 At that time, Habana had two products, Gaudi AI Training Processor and Goya AI Inference Processor.  It is worth noting that the Gaudi artificial intelligence training processor is already providing samples for specific hyperscale customers, and the Goya artificial intelligence inference processor is commercially available.

 Among them, Gaudi is the micro-architecture used by Habana to accelerate training.  Designed using the TSMC 16-nanometer process, the chip integrates eight TCP and GEMM engines in a cluster.  Goya is Habana's microarchitecture for accelerating inference.  Goya is manufactured on TSMC's 16-nanometer process, which is actually a simplified version of Gaudi.

 However, long before the acquisition of Habana, in August 2016, Intel acquired deep learning technology startup Nervana for $ 408 million.

 The main product of this company is the Nervana neural network chip. Its architecture design is very special. For the characteristics of AI requiring high-performance memory, it abandoned the standard cache system and switched to software to manage the memory system.

 Unlike Habana, however, Nervana does not have mass production products.  Three years after the acquisition of Nervana, Intel launched Nervana NNP-T and Nervana NNP-I.

 The Nervana NNP-T series, codenamed Spring Crest, is mainly for AI training. It is produced using TSMC's 16nm process and has a core area of ​​680mm2. It integrates 27 billion transistors and is equipped with 32GB of HBM2 memory. The frequency is 1.1GHz. TDP 150-250W.  Tesla series GPU acceleration chip.

 In contrast, the Nervana NNP-I series AI chip is much smaller, codenamed Spring Hill, which is mainly oriented to AI inference applications. The CPU part is Intel's 10nm process Ice Lake core, which consumes between 10-50W and has M.  2 and PCIe specifications, more compact and flexible.

 2. Who is more reliable?
 Just one month before the announcement of Habana, Intel reiterated its delivery plan for Nervana chips.  At that time, some analysts believed that the acquisition may indicate that customers were not satisfied with Nervana hardware and were unwilling to use Nervana again.  Since Nervana cannot keep up with the rapidly growing market, Intel will continue to explore other options.

 Now, Intel stopped TSMC's 16nm foundry Nervana chip, which also confirms this speculation.  Intel said it made the decision after seeking feedback from its engineers and major customers.  Feedback indicates that the second-generation Nervana designs, codenamed Spring Hill and Spring Crest, simply do not meet the requirements of those high-performance workloads.

 These customers also mentioned that Habana is a preferred platform to compete with Intel.  Among them, Facebook has made it clear to Intel: "You need a better chip."

 First, Habana network technology is likely to be one of the key reasons for Intel's decision to abandon Nervana in favor of Habana technology.

 Nervana's Neural Network Processor (NNP-T) scales using a proprietary interconnect, while Habana's Gaudi can scale to thousands of nodes over a standard 100Gb Ethernet.  Moreover, Gaudi even supports remote direct memory access RDMA, a feature that enables software to access memory in the entire fabric without increasing the burden on the remote CPU.

 This structure can significantly improve the performance of training large-scale neural network models to cope with the trend of doubling the size of neural network models every 3 and a half months, handling the trend of increasingly complex AI tasks.

 In addition, some analysts pointed out that full consideration of software and hardware coordination when designing the underlying compiler and software architecture also helped Habana chips achieve better scalability.  According to official figures, its distributed overall performance can be close to linear even when the number of processors is greater than 600. Compared to the Nvidia V100 GPU, the training performance is improved by nearly 4 times, which is a very remarkable result.

 Finally, Habana chips are not only more powerful, they have been shipping since the end of 2018.  In contrast, Nervana not only took three years from acquisition to product delivery, but also the product was repeatedly delayed.

 Buying out a competitor is a logical move by Intel. The success of Habana does not mean that he will not be replaced one day.  After all, Groq is doing its best, and more companies are bringing their chips to market.

 For Intel, as long as he sees the opportunity to surpass Nvidia, he will not hesitate. The key is whether he has selected the correct chip.

Thursday, 9 January 2020

compete with HuaBei? MeiTuan want a piece of cake of online payment market

Recently, Meituan officially launched the "cost-of-living · pay-to-pay" service, which is a product similar to HuaBei.



 From the screenshots posted by netizens, the "cost of living and pay" service was previously only available for some high-quality customers, and it has now been fully launched. Users who have enabled this function need only select "pay" as  Payment method, and then repay according to the bill before the repayment date.  The amount varies according to individual account qualifications, including 500 yuan, 1,000 yuan, and 1500 yuan, respectively.



 It is reported that the User can pay back by the same month I'm, or pay back next month, there is no fees.  Payment is made on the 1st of each month, and repayment is made on the 8th of each month. Online payments are available using the Meituan App, including Meituan, Reviews, Meituan Takeaway, Meituan Taxi, Meituan Shopping and other apps.

 It should be noted that the purchase order of Meituan is a financial product of Meituan, which can be used to enjoy the payment experience of "consumption first, then payment", and points can be exchanged for vouchers. Normal use does not check credit.

 Regarding the above-mentioned behavior of Meituan, some people in the industry believe that the current limit and scale of the "buy orders" are not large, and the thickness of the bills is not enough. Under the current market environment, the benefits can not cover the costs at all.  There will be big gains.

 Another view is that Meituan's move is to imitate Huayan. With the successful case of Huayuan, Meituan's "pay" products will not easily fail, and the risk of small quota may be smaller.

Most of our user are interested in this post:

可控核聚变:EAST 亿度千秒突破与商业化路径

可控核聚变:EAST   亿度千秒突破与商业化路径 关键词 :  #可控核聚变  #EAST  #托卡马克  #BEST  #聚变商业化   #清洁能源   1.  技术原理 1.1  核心技术概述 可控核聚变通过让轻原子核(主要是氘  D  与...