At 12:12 on 27 July 2022, the ZK-1A launch vehicle, developed by the Institute of Mechanics , Chinese Academy of Sciences (hereinafter referred to as the Institute of Mechanics of the CAS), was successfully launched from the Jiuquan Satellite Launch Center, successfully placing six satellites into the scheduled orbit by means of "one arrow and six stars", making the first mission of the ZK-1A launch vehicle a complete success. As the preferred choice for small and medium-sized satellite launches, it has enriched the spectrum of China's solid launch vehicle launch capability. V Fund invested in CAS SPACE in April 2019 as an early shareholder of the company and is confident that the company will become a leader in commercial spaceflight in China. CAS SPACE "Supported by the 14th Five-Year Plan of the Chinese Academy of Sciences (CAS), the "ZK-1A" launch vehicle is another innovative attempt of CAS in the field of space science and technology. Facing the development needs of space science and space technology, guided by the idea of "engineering science", with innovative, advanced and efficient design ideas, the development of innovative, advanced and economical launch vehicles is of great significance to promote the change and innovation of China's launch technology and development mode, and to promote the development of space science. The "ZK-1A" launch vehicle, the first solid launch vehicle developed by the Institute of Mechanics, Chinese Academy of Sciences with the participation of CAS SPACE, is the largest solid launch vehicle in China at present, and has entered the advanced ranks of the world's solid launch vehicles in terms of carrying capacity, accuracy in orbit, design reliability and cost performance. The rocket is a four-stage solid launch vehicle with a take-off weight of 135 tons, a take-off thrust of 200 tons, an overall length of 30 meters, a core stage diameter of 2.65 meters, a fairing diameter of 2.65 meters in the first flight state and a 500km sun-synchronous orbit capacity of 1,500kg. It is reported that during the development of the "ZK-1A" launch vehicle, the relevant institutes of the Chinese Academy of Sciences and the Fourth Academy of the Aerospace Science and Technology Corporation jointly launched a scientific research mission, adhering to the design concept of "innovation for limitless times". The research team has made breakthroughs in a number of key technologies, including the overall optimized design and test technology of large-tonnage solid rocket, advanced power system and thrust vector control technology, centralized-distributed modern aerospace electronics technology, low-cost arrow structure and separation technology, intelligent flight control technology, ground use of large-tonnage solid rocket and thermal launch technology. The innovation of the research paradigm has enabled high-performance and low-cost agile development, making an important contribution to the technological change in the field of China's launch vehicles. The "ZK-1A" launch vehicle has a series of outstanding advantages, such as large launch capacity, high accuracy in orbit, high inherent reliability, fast response time, high launch efficiency, low security requirements, low launch cost, flexible and convenient use, good environmental adaptability, etc. It is suitable for medium/low orbit low-cost rapid network launch of medium/small satellite payloads. The mission will carry six satellites, namely the New Space Technology Test Satellite, the Orbital Atmospheric Density Detection Test Satellite, the Low Orbit Quantum Key Distribution Test Satellite, the Electromagnetic Assembly Test Double Star by Innovation Academy for Microsatellites of Chinese Academy of Sciences and the Nan Yue Science Satellite developed by the Shanghai SastSpace Technology Co., Ltd. The New Space Technology Test Satellite is a versatile, scalable and fast-responding new space technology test satellite platform that will carry out experiments on new space science payloads.

There is a real opportunity to change an industry, even if it costs whatever you give, try it. If a non-industry startup tells you that he wants to make a GPU chip to catch up with Nvidia, but has only prepared a 5-page PPT so far, how would you, as an investor, choose? Zhang Wen is one such non-industry entrepreneur. He has a JD from Harvard University and extensive management experience. In his early years, he was a senior investor in the Pan-Asian market on Wall Street; in 2011, Zhang Rujing, the founder of SMIC, set up Enraytek Optoelectronics Co., Ltd., and Zhang Wen was invited to be the CEO of the company; in 2018, Zhang Wen became the president of SENSETIME and led the process of setting up its headquarters in Shanghai. However, when he decided to found BIREN TECH in 2019, Zhang Wen still seemed to the outside world to be a layman in the GPU (graphics processor) field with no industry background. On 16 May 2019, the United States included Huawei in its list of entities and several US chips were cut off to Huawei. As a result of this, the first small climax of chip startups was set off in the country. The hope for entrepreneurship was also lit in Zhang Wen's heart, who saw that there was a huge domestic market for AI chips, but more than 90% relied on foreign products. The AI chip has its own unique opportunities in China: firstly, because the theoretical architecture of AI chips is also being broken through; secondly, because of China's data advantage, AI chips need scenarios and data iteration the most. GPUs refer to graphics processors, which are microprocessors that specialize in graphics-related work. In the post-Moore era, GPUs are increasingly used for applications beyond graphics rendering and are used in general-purpose computing. In particular, a number of startups specializing in GPUs have been created as artificial intelligence applications have become more popular. However, the GPU is a tough nut to crack, even with the chance to survive in Nvidia's empire. All around him, people voted against Zhang Wen's new dream. From the standpoint of an investor, he himself was against it: Don’t vote. "At that time in the big chip track, I was the only founder who didn't know anything about technology, and I didn't even know one of the big names in GPU making at that time." Zhang Wen told early investors afterwards, "If I were you guys back then, I wouldn't dare to invest in Zhang Wen." Things always turn out unexpectedly, and not only has BIREN TECH raised money, it has set a record. From its incorporation on 9 September 2019 to the completion of its Series B round of funding in March 2021, in just 18 months, BIREN TECH raised more than RMB 4.7 billion. Investors have made BIREN TECH to set a funding record for a domestic chip startup. Zhou Zhifeng, a partner at early stage investor QiMing Venture Partners, was certain: when we met Zhang Wen and the founding team of BIREN TECH, they had the right people, the right ideas, and the right market needs to capture. On the evening of March 31, 2022, BIREN TECH proved Zhou Zhifeng's judgment: its first general-purpose GPU chip, the BR100 series, lit up successfully. Due to Shanghai's quarantine, Zhang Wen and nearly 100 core employees, who spent more than 60 days together at the company, witnessed the first milestone for BIREN TECH. For Zhang Wen, this is just a new starting point for BIREN TECH: "From lighting up the flow to entering mass production, the team went all out in just two to three months, and our co-CEO Li Xinrong, who was previously the global vice president of AMD, told me that in his 35-year-long experience in the chip industry, a general-purpose GPU chip with such a large arithmetic power and such a high degree of difficulty was lit up and smoothly produced in one-go mass production, this speed is very rare in the entire chip industry." In an industry with wide tracks, high barriers and long cycles, there is so much uncertainty. In the vast process of technological development, there are big technology rising stars whose myths break and fall; there are also startups that survive the winter and revived. "One investor once told me, 'I absolutely believe you are doing this with extreme passion, even if you lose all your money this time, I will invest in you next time. 'This is a very important statement to me. There is only one path to the mountain, so I have to make this work." Zhang Wen is ambitious. "An entrepreneur is someone who dares to dream the dreams, "he writes on his WeChat signature file. With Zhang Wen, BIREN TECH is embarking on a daring dream journey. No.1 headhunter Although he doesn't understand the specific technical details, Zhang Wen knows where his strengths lie: "My previous experience in entrepreneurship is a good testament to my ability to integrate capital, talent and resources." Since he didn't understand it, he hired someone to do it. He approached a friend at Harvard, who gave him a list of GPU Big Names. The friend dictated that he would be successful if he could bring together half of them; after all, it was hard to find the right talent from China. Of the two major international GPU companies, Nvidia has no core R&D team in China, and AMD has only a small R&D team in Shanghai. The only ones who have started from the GPU architecture level in China are the original Zhaoxin team and the current Hisilicon team. The first "Big Name" that Zhang Wen targeted was Hong Zhou, who was responsible for the self-research GPU team from Hisilicon. Hongzhou and Zhang Wen first met at the World Artificial Intelligence Conference 2019 in Shanghai, perhaps due to a subtle change in the general environment, and Hongzhou was looking to try something different, so the two hit it off immediately. After meeting him twice, Hong Zhou agreed to take on the role of CTO of BIREN TECH. Zhang Wen said: "Startups can't afford to pay an annual salary of ten million, if you want to leave then go early, do not do it halfway and leave." Hongzhou replied, "For money, I probably wouldn't go into business that way, I just want to make this work." At the end of 2019, Hongzhou said goodbye to his family and took his last flight back to China before the pandemic outbreak. To prevent the spread of the epidemic, he consciously wore two extra plastic bags on his body. For two years, the CTO arrived at the office at 7:30 am and left at 10:00 pm, day after day. In addition to Hong Zhou, BIREN TECH quickly gathered a group of technical talents, including Jiao Guofang, who founded Qualcomm's Snapdragon GPU team and led the development of the 5th generation Adreno GPU architecture; Xu Lingjie, former head of AI infrastructure at Aliyun; Zhang Linglan, former VP of overseas GPU division at HYGON; Tang Shan, former head of AI Lab at EDA software giant Synopsys; Li Xinrong, former global VP and GM of China R&D Center at AMD; and Yang Chaoyuan, former GM of Shanghai at Nvidia. When it comes to the speed of finding talent, Zhang Wen is satisfied with his achievement: "They all call me No.1 headhunter in China." Not only do we have the core technology Big Names, but more importantly, we have built a "complete" team. "Most of the team are former directors and vice presidents of large factories, and the team is formed by the master with apprentices, confined to a small circle, especially prone to shortcomings. But to do 'big chip', from architecture design to hardware, SoC, to engineering, software ecology, to financing, recruitment, etc., there can be no shortcomings, and at that time we saw that the team formed by Zhang Wen was the most complete." Zhou Zhifeng said. Once the team was ready, it was followed by the second task: funding. The investors asked Zhang Wen if 100 million dollars was enough. But when Zhang Wen opened his mouth, he said an astronomical figure - 10 times higher: 1 billion dollars. He calculated that the starting price of a chip company is about 200 million dollars, and after the flow of chips is done, a software team of thousands of people will be needed, which is also an expense measured in billions of dollars. If they want to make the ecology complete, 1 billion dollars is only a small amount. Zhou Zhifeng was not shocked by this figure, instead he agreed with Zhang Wen's idea: the ratio of hardware and software engineers in "big chip" companies is usually 1:3, or even 1:5, but the ratio of hardware and software engineers in most other chip companies is 1:0.5, or even 1:0.2. "Chips require three elements: capital-intensive, talent-intensive, plus resource-intensive, is a very expensive industry; second, from the product landing to slowly generate cash flow, at least two to three years; finally, from my past experience in entrepreneurship, startups must allow trial and error, trial and error also requires costs." These three points make Zhang Wen clearly aware that BIREN TECH has to "accumulate as much as possible". It was a big gamble, and Zhang Wen had his moments of hesitation: should we make a modest product and go from there, or should we gamble big and go for a product that would really change the industry? Despite the fact that the latter option might be a faliure, the investors and Zhang Wen were on the same page: there was a real chance to change an industry, even if it costs whatever you give, so they had to try. A Highly Dedicated Entrepreneur With more money comes controversy. The outside world has labelled Zhang Wen as a capital maniac, and there have been criticisms of the "sky-high" financing of BIREN TECH. Indeed, the overheated GPU track has seen new startups opening their mouths with valuations of multi-ten billion dollars at every turn. This is somewhat similar to the controversy surrounding the development of BOE, which has been expanding at a breakneck pace since it entered the panel industry in 2002 with debt-based operations and leveraged financing for mergers and acquisitions, raising over 70 billion yuan to date, almost the largest financing an industrial company can do in the capital markets. Some media have commented that it is extremely skillful in "capturing money" and that its overall level of capital operation is "excellent" in the industry. "Being open within yourself may give you more power. Because I had a lot of people who didn't understand from the first day I started the business." At BIREN TECH, Zhang Wen insists on being paid just one dollar, and he doesn't want to entangle his personal interests with those of the company during the start-up process. In the early days he would look at the reviews, and now he doesn't care about the misconceptions. He is clear that whether the reviews are complimentary or questionable, the most important thing is to get the product made. On October 8, 2021, BIREN TECH gave an advance warning that the BR100 series will be "available to the market next year". The BR100 series of general-purpose computing products will focus on a wide range of computing application scenarios, including artificial intelligence training and reasoning, and general-purpose computing. Promises were fulfilled.On 28 March 2022, during the closure of the company due to the epidemic, Zhang Wen and nearly 100 core employees chose to stay together at the company to usher in the most critical point of the company. Three days later, on the evening of 31 March, the BR100 series, the first general-purpose GPU chip from BIREN TECH, was successfully lit up. In the monumental group photo, the engineers could not hide their excitement. The BR100 series of chips, according to BIREN TECH are based on their original chip architecture, using a 7nm process and incorporating a number of cutting-edge chip design, manufacturing and packaging technologies, including Chiplet (core particle technology). The BR100 series will be the representative of the domestic chip industry at HOT CHIPS in August this year, where it will compete with Nvidia, AMD and Intel, the three international giants of the chip industry, and is a key benchmark for the global high-performance chip industry, especially for original chip architectures. Zhang Wen is somewhat paranoid, he is extremely demanding of this chip and wants it to be in the same league as the top manufacturers. This is perhaps a habit that has been cultivated since childhood. "When I was a child, my mother once said to me, 'Do everything carefully and to the best of your ability, even wipe the table to get it the cleanest, or you don't do it'." Zhang Wen said. Since then, he feels that to do something either don't do it, or do it to the best of his ability. After returning to China in 2011, Zhang Wen stayed in the deserted Shanghai Lingang district for four years in order to raise the yield of LED chips from 70% to over 95%. Before joining the company, Zhang Rujing asked him if he had any requirements. He said that it was enough for him to just drink Starbucks. Until he left Lingang, this wish did not come true. "People who can succeed in entrepreneurship are basically a bit paranoid." Zhang Wen doesn't shy away from this statement, "I have a logical mindset that is different from others, if 90% of people agree with something and say to do it, I won't do it, if it's 90% of people who are against it and say don't do it, I will do it instead." The choice of GPU track also followed this logic. However, Zhang Wen's choice stirred up a thousand waves. 2020, former Nvidia global vice president Zhang Jianzhong founded Moore Threads, which raised multi-billion dollars in 100 days; almost at the same time, Metax-tech was established, with founder Chen Weiliang as a former AMD director, and Metax-tech also received four rounds of funding within a year. There is a growing consensus that there is more space in this track. According to IDC data, the global chip market will reach 35.2 billion dollars in 2022, with GPUs accounting for the most rapid growth and expected to account for 57% by 2025. In addition, IDC expects the Chinese market to reach 525.2 million server shipments in 2025, with a market size of 35 billion dollars. "This is a market of multi-hundred billion dollars, in fact all our current friends are not considered competitors, we should love this industry together and not get in ‘involution’" .Zhang Wen said that at present, BIREN TECH has already got many orders from important customers. On November 15, 1971, Intel wrote the bold words for a chip the size of a fingernail cap: to start a new era of integrated circuits. That year, Intel defined the CPU, and in 1999, Nvidia defined the GPU. now, the story of Chinese chip companies has just begun. ​

Recently, V Fund investee company "EMOTIBOT" has completed D1 financing round of nearly 100 million yuan,which was jointly invested by KYMCO Capital, Jiangsu Cultural Investment and JUNCI Investment. V Fund is the investor of EMOTIBOT's C financing round. Since V Fund's investment, the company has completed several financing rounds such as Series C+ and Series D respectively, and continues to be recognized by head investment institutions. EMOTIBOT Founded in 2015 by Jian Renxian, former Vice President of Microsoft (Asia) Internet Engineering Institute, EMOTIBOT has been working on natural language understanding for many years, with knowledge, interaction, emotion and machine learning as its core. Starting from the B-side, where the need is more urgent, EMOTIBOT systematically trains smarter digital employees by following the path of "general technology base layer → industry terminology and scenario know-how → autonomous operation and maintenance without code", which is similar to real-person staff. In a market that is now generally focused on replacing low-end labor, EMOTIBOT goes a step further by providing assisted intelligence for high-end white-collar workers and helping enterprises integrate and revitalize their knowledge assets through a shared brain of multi-scene and multi-module digital employees to achieve a full-scene intelligent transformation. As a head player focusing on the NLP track in China, Jian Renxian, founder and CEO of EMOTIBOT, said: "We have always believed that knowledge asset-based management will become the new generation of operating system for enterprises, and we also hope that every customer can have their own digital employees and AI assistants, and NLP technology will be a strong support to achieve the above vision. Under the path of continuous iteration of digital employees, as long as there is a continuous input of knowledge assets, robots can learn endlessly; as long as the needs are clearly sorted out, any language-related scenario in any industry can have its own 'EMOTIBOT digital employee'; and through cooperation with ecological partners, using NLP to empower RPA, By collaborating with ecological partners and using NLP to empower RPA, OCR, CRM and other technologies and vertical applications in various fields, multi-ten thousand scenarios will be created. In the future, EMOTIBOT will continue to promote various in-depth cooperation with strategic investors and other eco-partners to continuously broaden the boundaries of product applications." EMOTIBOT has developed its own NLP technology to bridge short and long texts, combined with deep learning technology to build a composite AI general technology base layer with logic-based symbolic AI and data-based neural AI, building a strong technical barrier. NLP is the core technology that gives AI its cognitive intelligence, and the degree of generality of NLP technology determines the extent to which digital employees can be produced at scale. From a linguistic perspective, language is divided into long text and short text, which belong to two language systems: written text and conversational oral language, and their corresponding main business application scenarios are knowledge and interaction respectively. A good NLP technology should be able to handle both long and short texts, but players in the market currently only focus on one of the two, mainly because the technology relied on to achieve natural language understanding is still relatively traditional. Short text interaction vendors usually use open source algorithms, rules, keywords, standard Q&A, etc., and rely on relatively limited corpus annotation in fixed scenarios to achieve the effect of barely replacing humans, but cannot achieve more intelligent dialogue effects such as multi-round Q&A and intent recognition; while long text is not practical to commercialize with pure machine learning due to the long length of the document and the large knowledge level. The depth of the corresponding application vendors can usually only handle documents with relatively uniform formatting and corpus in specific business scenarios, and cannot parse complex sentences or perform automatic annotation, and is even more powerless for spoken interaction scenarios with low standardization such as inversion, omission and multiple rounds. The lack of technical depth determines the low level of product versatility, and the switching of industries and scenarios will rely on customization, which will easily expose the lack of momentum in productization, engineering and scale implementation. Deeply aware of this, EMOTIBOT chose a different technology path from that of a purely long-text application developer or short-text intelligent interaction vendor, using generic and composite NLP technology to bridge knowledge and interaction and build a complete technology stack. On the one hand, we have hired professional linguists to annotate the corpus at the beginning of technology development, and have developed a variety of technologies to deal with complex, non-standard and difficult sentences, including person recognition, denotation disambiguation, ambiguity understanding, entity linking and recognition, semantic role analysis, dependent syntax analysis, sentence vector contextual denotation analysis, error-tolerant recognition, etc. We have accumulated more than 30 NLP capability models and over 5 million linguist annotated. On the other hand, it combines deep learning technology to build a composite AI with logic-based symbolic AI and data-based neural AI, fusing linguistics and neural networks to map knowledge, opening the black box of AI thinking while also providing technical support for model compression and light annotation volume at the same accuracy rate. Through 100% self-research, EMOTIBOT has built a solid technical base with natural language processing, knowledge engineering, multimodal emotional computing, deep learning, intelligent voice technology and text data middleware to equip digital employees with the knowledge base, communication, interpersonal and continuous learning capabilities. The funding will be used to promote standardized SaaS products, continue to consolidate the technical barriers and product advantages, and help more head and waist customers to achieve intelligent transformation of their enterprises.

Recently, V Fund's early stage investment company, Horizon Robotics, announced that it has received a strategic investment from China FAW Group Corporation (FAW Group), which will be used to strengthen the forward-looking technology research and development of vehicle-grade AI chips as well as the construction of engineering implementation capabilities. V Fund has been in contact with Horizon Robotics since 2017, and participated in its A3 and B2 financing rounds. Horizon Robotics Founded in 2015, Horizon Robotics provides empowering services with ultimate performance, openness and simplicity to the industry, with edge AI chips as its core. Since its establishment, Horizon Robotics has received strategic investments from many automotive companies such as FAW Group, GAC Capital, GWM Group, Dongfeng Management Capital and BYD, as well as many upstream and downstream companies in the industry chain such as Intel, SK Hynix, CATL, XINGYU CO., Ltd, WILLSEMI and sunnyoptical, etc. Its technology concept and R&D achievements are highly recognized by the industry. Vehicle-grade AI chips are becoming the "digital engine" of smart cars. Based on the technology concept of combining hardware and software and collaborative optimization, Horizon Robotics is committed to creating an efficient bottom open technology platform with the core of "chip + algorithm + tool chain" through technological breakthroughs in key aspects such as AI computing architecture, deep learning-based autonomous driving technology and software 2.0 infrastructure to comprehensively improve the efficiency of technology development and application iteration. For L2~L4 full-scope vehicle intelligence application scenarios, Horizon Robotics has driven the rapid iteration of the Journey chip with forward-looking technology, continuing to lead the innovation breakthrough of domestic vehicle-class AI chips. Horizon Robotics' third-generation automotive-grade product, Journey®5, combines high performance and large computing power to meet the needs of high-grade autonomous driving mass production applications. Journey®5 has won the cooperation of many mainstream car companies such as BYD, NIUTRON and FAW Hongqi. It is the only two in the world and the first one in China that can be mass-produced with 100 TOPS-class computing power AI chip. Based on the concept of open and win-win cooperation, Horizon Robotics' independent innovative technologies, breakthrough products and solutions have brought new business value to a wide range of industrial partners. By the end of 2021, the cumulative shipment of Horizon Robotics Journey chips has exceeded 1 million units. Currently, Horizon Robotics has signed pre-installation mass production spot projects for over 70 models with more than 20 car companies, making it the largest Chinese automotive-grade AI chip company in terms of mass production scale. As an innovation leader in China's automotive industry, FAW Group and Horizon Robotics have long established a deep partnership. Since 2020, Horizon Robotics has reached strategic cooperation with FAW Nanjing and FAW Intelligent Network Development Institute, and in May 2022, Horizon Robotics was awarded the application of FAW Hongqi's new model project, and will work together with industry partners based on multiple Journey 5 chips to help FAW Hongqi build a high-level autonomous driving solutions. This strategic investment further deepens the cooperation between the two parties. In the future, Horizon Robotics will continue to work together with FAW Group and more industrial partners to create innovative and efficient development paths based on their respective core R&D capabilities, and promote China's smart car industry to achieve global leadership.