Another AI unicorn was born.

This week, Xiaobing, which was spun off from Microsoft for a year, announced the completion of round A financing, led by Hillhouse. Although the amount of the financing was not disclosed, the investment community knows that Xiao Bing's valuation has exceeded that of a unicorn ($1 billion). Through this heavy investment in AI, outsiders are beginning to discover:The color of Hillhouse hard science and technology is getting stronger and stronger.。

Zhang Lei, founder of Hillhouse, did not make a public appearance at the 2021 World artificial Intelligence Conference in Shanghai a few days ago, but participated in the discussion of the Strategic expert Advisory Committee.Hillhouse artificial Intelligence ProjectIncludeDeep learning framework "first-class technology", open source framework for vector search "Zilliz", AI computing resource pooling software "trend technology", as well as AI chip enterprises "horizon" and "wall technology"". Hillhouse judged that "perceptual intelligence has made a transition to higher-level decision-making intelligence, and a large number of interdisciplinary breakthroughs are being realized."

Combing the sales in the first half of the year, we can find that Hillhouse's hard technology investment so far this year can be described as a chain layout. In June, it invested in Robotics, Wesai Technologies, a lidar company ($300m with Meituan Xiaomi), and in May, hundreds of millions of people invested in the new energy network Star charging Pole and DPU chip Nebula Zhaopin. Looking ahead, Jifei Technology, Mimo Zhixing, Guoyi Quantum, and Weiger belong to the fields of UAV, autopilot, quantum precision measurement and quantum computing, and low code. According to incomplete statistics, Hillhouse made more than 80 shots in the hard technology field in the first half of the year.

Looking at the VC/PE circle, Hillhouse has always been good atEcological investmentThe most typical case is that Hillhouse was previously inMedical fieldThe layout. Now in the field of hard technology, Hillhouse is clearly replicating what it does best.

Exclusive disassembly of Hillhouse chip investment logic: grasping "big chips" and gamblers

What is little known is that Hillhouse has quietly created one.Chip semiconductorInvestment territory.

Core Yaohui, Core Huazhang, Horizon, Skyline Hedar, Star Semiconductor, financier Biji Technology and Nebula ZhaopinCorresponding to the IP and EDA designs at the top of the chip industry chain, to the vehicle, dynamometer, and extremely difficult mobile phone baseband in the application chip, to general-purpose GPU and DPU.

Hillhouse has invested in these, and almost all of them are leaders.

"long snow trails and lots of opportunities" is Hillhouse's definition of the chip industry, and the last time it was said was that they invested in Dacheng's biomedicine. So, is Hillhouse casting a net?

The investment community obtained a speech from Hillhouse Technology Investment, which revealed that Hillhouse has at least two selection criteria: first, only look at "big chips", whether CPU, GPU, or vehicle power semiconductors, are subdivided tracks with huge market scale and extremely high ceilings. On this basis, the demand put forward by Hillhouse, whether it is large capital or ultra-long-term, is willing to be highly satisfied.

Here is a typical example-- wall technology for making general-purpose smart chips. In August 2020, Gao Gao led the investment in the 2 billion Pre-B round of financing. In the view of many semiconductor industry FA, Hillhouse's entry at that time "has a strong signal meaning". Since then, it has raised a total of 4.7 billion yuan by March this year, becoming the super unicorn with the fastest development and the largest financing scale in the industry.

Sort out the layout of large chipsThe key to Hillhouse's selection of a team can be summed up as "finding a strong man".. For the chip semiconductor industry, there may be two strong standards: first, enough experience + enough influence. If it is changed to some kind of convertible market standard, it can be said that it depends on whether the founder can respond with one voice or not. According to this dimensionXia Lusheng of Xing Si, Zhang Wen of Bi Jing, Wang Li Bing of Xin Hua ZhangThey are all typical strong men.

Second, the team must pursue technological leadership without distractions, which is the core competitiveness of Changxue Road and is highlighted in Hillhouse's internal speech.

Hillhouse has a principle--"I don't know how to vote for a technical follower.". If you imitate other people's technology, even if you achieve the first position in the subdivision of the industry in a short period of time, the high probability is not a good match for Hillhouse.

Layout autopilot: how does Hillhouse vote?

In fact, Hillhouse is right.Autopilot layoutIt also originated from the chip. As early as 2015, Zhang Lei invested in the Angel Wheel.HorizonLast month, Horizon successfully rolled out the 5 chip, becoming the only intelligent chip provider in the industry to cover the whole scene from L2 to L4.

Yu Kai, founder of Horizon, recalled that the company was not favored by the industry at the beginning of its establishment, and the two major barriers of technology and capital included in AI and chips "deterred the vast majority of investors." But Zhang Lei not only supported you from the Angel Wheel, but also encouraged Yu Kai: "entrepreneurs should enjoy a period of time when they are not understood and others think you are a fool." "

At the beginning of this yearHillhouse, together with Wuyuan Capital and Today Capital, once again led the horizon's C7 round of financing.。

In addition to the on-board chip, there are many key points in the research and development and commercial landing of autopilot. Before that, Hillhouse investors have concluded: "Hillhouse's investment in autopilot is first of all around the core nodes in the industrial chain-including autopilot systems, autopilot chips, lidar of the most core sensors, and autopilot computing platforms. Find the most critical link, and then select the strongest team in the subdivision direction to support them to develop the relevant direction. "

Take the autopilot system as an example, in the process of integrating hardware and software, we need a perfect "Windows system", which is also the most important underlying technology except the chip. In this segment, Hillhouse invested in AutoCore.ai with Foxconn. The company has built the first generation of autopilot open development platform with TierIV of Japan, and has established deep cooperation with global leading manufacturers in chip, Tier1 and OEM fields and has completed many project deliveries.

LidarAsPerceptual hardware necessary for autopilotIt is more favored by capital.Wo Sai TechnologyIPO ended in March this year, but still could not stop the enthusiasm of investment institutions. In June, the unicorn quickly completed a financing of more than $300m from D-round, led by Hillhouse.

Looking back on the first half of the year, we also found that Hillhouse has invested heavily in a number of L1-L3-level autopilot service providers, and the logic behind it is not difficult to understand. Hillhouse believes that "the landing speed of autopilot, especially L1-L3, is much faster than expected, so there will be rapid commercial progress in logistics and distribution, food delivery, cleaning, agriculture and other areas."

At present, the L3-level product "Little Magic Box" invested by Hillhouse has been put into the market and has become a supplier of logistics unmanned vehicles for many large companies. Jifei Technology can also be regarded as one. Its main business, such as drones and auxiliary driving, are mainly used in farmland scenes. In addition, investors in Acer's round A financing of nearly 100 million yuan also include Hillhouse, which is also a rare full-stack self-driving service provider with complete self-research capability of software and hardware in the industry, with products covering many levels of L1-L4.

Quietly cast 80 Hillhouse in half a year. Why do you believe that hard technology will usher in the tipping point?

In recent yearsMany outsiders' impression of Hillhouse still lingers on investing in Tencent, JD.com, holding Belle, and the ecological layout of health care.. But if the investment portfolio of Hillhouse in the past three yearsIt can be found that this investment giant is taking up a higher and higher proportion of its investment in hard-tech companies.. In fact, technology has always been a main line of Hillhouse investment.

In early 2021, Hillhouse Ventures released a surprising set of figures: it invests more than 200 projects a year, with technology-driven companies accounting for 78%.Among them, there are more than 80 investments in hard technology. But it took Hillhouse half the time to break the number of "80" this year.。

Chips, autopilot, robots and artificial intelligence are all hot directions.To some extent, it is also the place that big funds must compete for. Zhang Lei has always stressed that it is strange to be upright, so what is the "surprise" in Hillhouse's hard technology investment?

The quantum could be a. On more than one occasion, Hillhouse mentioned a company called Guoyi Quantum, whose strategy is to use quantum precision measurement and quantum computing to cut the point of domestic high-end instruments in the "neck".

And basic software. Li Qiang, a partner of Hillhouse Venture Capital, said in a speech at Zhiyuan Conference in Beijing that Hillhouse Venture Capital had systematically laid out a large number of early-stage companies "in the fields of IT, DT, AI, computing and storage, deep learning framework, security and operation and maintenance, open source, etc." For example, in terms of open source, Hillhouse has invested in Zilliz, EMQ Yingyun Technology, PerfMa, first-class technology and other companies.

"Infra projects are now given priority to Hillhouse," said a senior FA in To B industry. There are a lot of investors in this area, but the Hillhouse approach is the most systematic.

"most investment institutions look at the Infra team of only one or two people, at most a small Team, but Gao Zhi only looks at the team in this direction is relatively large. So they are not dotted to vote, there must be a road book, we can see that they are looking for the layout step by step according to the clues. "

According to data from inside Hillhouse, Hillhouse made it clear"be optimistic about the four subdivisions of semiconductors, cutting-edge technology, new energy and intelligent hardware in the field of science and technology in the next two to five years."And set theHard science and technologyCalled"the historic window of structural investment」。

There are many factors behind this, among which, affected by the environment, the accelerated return of overseas hard-tech talents is a key indicator, which greatly complements the technological strength of Chinese companies in a short period of time. Hillhouse believes that hard technology entrepreneurship is in the process of rapid evolution of high demand, high power, high barriers and high-level teams, which is the premise of Hillhouse betting on the technology track-these factors ensure that enterprises promote industrial progress at the same time. You can also find a balance between technology and business.

So where does Hillhouse's "technological obsession" come from? In May this year, Zhang Lei mentioned that becauseHillhouse sees a great opportunity brought about by the change in the paradigm of science and technology., includingInfrastructure software, artificial intelligence, autopilot, life sciences, immune genes, precision therapy, as well as new materials, environmental science, carbon neutralization, advanced manufacturingAnd so on, it can be said that the whole scientific innovation, the outbreak of new species of scientific and technological innovation.

This is reminiscent of Hillhouse's investment in Baiji China. In 2014, Hillhouse invested in Baiji Shenzhou A round, followed by eight consecutive rounds of bets. At the time of the first move, few people believed that China could make its own innovative drugs. But Hillhouse's investment requirement is to find something that will make a big difference in the next five to 10 years before the consensus occurs.

Today, such a familiar scene is repeated in the Hillhouse hard technology landscape.