CCTV News:The National Health and Wellness Commission, the Ministry of Housing and Urban-Rural Development and other departments issued a notice on printing and distributing the Measures for the Administration of Family Nurseries (for Trial Implementation), which clarified the provisions on the number of children in family nurseries and the per capita construction area of infants and young children.

According to the notice, a family nursery refers to a place that provides full day care, half day care, time care, temporary care and other care services for infants under 3 years old by using houses, and the number of children in each family nursery must not exceed 5. Every 1 caregiver in the family nursery can care for up to 3 infants, and the per capita construction area of infants in the family nursery shall not be less than 9 square meters. The words "care" should be indicated in the name of the family care point, and the "family care service" should be defined in the service scope and business scope. If a profit-making family nursery is held, it shall apply to the local market supervision department for registration according to law, and the registration authority shall promptly push the registration information of the family nursery to the health department at the same level.

After the transformation of the old Jingyu line and Tongyun intersection, the traffic efficiency has been greatly improved. Photo courtesy of Tongzhou Highway Branch

　　On September 18th, it was learned from Tongzhou Highway Branch of the Municipal Communications Commission that the road network system of Beijing City Sub-center has recently ushered in a new round of optimization and upgrading. Through the implementation of a number of key projects, such as unblocking renovation, channelization of intersections and road widening, many long-standing traffic congestion nodes have been effectively managed. The morning rush hour traffic efficiency of some congested sections of the old Beijing-Yushu line has been improved by 45.8% at the maximum, and the queue length has been reduced by over 56%.

　　It is understood that the reconstruction project of the intersection of Jingyu old line and Tongyun Road was started on September 12, and it was completed in only 72 hours. Through refined design, some green separation zones were removed, and the east entrance was optimized into five-in, three-out and eight-lane configuration, so that the traffic capacity of the intersection was obviously improved.

　　The data shows that after the reconstruction, the morning rush hour traffic efficiency of the old Jing-Yu line has increased by 45.8%. The average speed of the old Jing-Yu line from east to west reached 31.5 kilometers per hour, an increase of 49.5% compared with the same period in 2024. It is worth mentioning that the average queue length of the east import morning peak has dropped from 486 meters to 213 meters, a drop of over 56%.

　　Recently, the traffic optimization project of Chunming West Road, where the east gate of Beijing School is located, was completed. As the main passage around the school, this road section has a high concentration of people and traffic during the daily school hours, and the traffic pressure is outstanding. In this reconstruction, the traffic order has been obviously improved by scientifically adjusting the layout of the central isolation belt, adding special U-turn lanes and updating traffic markings and warning signs simultaneously. The optimized road section effectively disperses the traffic flow during school hours and reduces the interweaving of people and vehicles.

　　At the same time, the widening project of Tonghuai Road (from Songzhuang South Street to Tongyan Expressway) is in full swing, striving to achieve the goal of opening to traffic at the end of October. The project has been progressing smoothly since it was started on September 6th. At present, 21% of the total engineering quantity has been completed in the civil works, and the relocation and renovation of communication and power pipelines have been carried out in an orderly and synchronous manner, with the completion rates reaching 7% and 10% respectively. The transformation focuses on the west side motor vehicle lane, which is widened from 2 lanes to 3 lanes, and an independent non-motor vehicle lane with a width of 4 meters is added simultaneously to realize the separation of vehicles and vehicles, and improve the road capacity and safety.

　　In addition, the dredging project of Liuhe bridge section of the old Beijing-Yulin line has now entered the preparatory stage of implementation. The road section is two-way and three-lane, which restricts the regional traffic flow for a long time. At present, the bidding plan for the project has been formulated and the bidding announcement will be issued soon.

　　As the east-west traffic artery in the north of the sub-center, after the completion of the reconstruction project of Liuhe Bridge, the regional traffic network will be further improved, the road capacity will be improved, and the overall traffic operation of the sub-center will be strongly supported.(Sun Hongyang)

　　related news

　　In the future, many roads will be built in the west of Science City.

　　On September 18th, it was learned from the Management Committee of Future Science City that many roads, such as Qixin Middle Street, Qixin South Street and Qiyan Road, have recently entered the site for construction, which will further penetrate the "capillary" of the road network in Zhuxinzhuang area and help the development of innovative industries in the region.

　　In this batch of road projects, Qiyan Road (Qixin North Street to Xiaoshahecun North Street) is scheduled to be completed by the end of 2026. The road is planned to be an urban secondary trunk road, which is divided into two sections: north and south. The south section (Beiqing Road to Qixin North Street) has been completed and opened to traffic. The north section is about 1.4 kilometers long, and municipal comprehensive pipelines are laid simultaneously. Qixin South Street (Zhuxinzhuang Middle Road to Agricultural College East Road) has been put into construction, with a road length of about 440 meters, which is expected to be completed before the end of 2025. The west section of Qixin Middle Street (from Dongfu Road of Beijing-Tibet Expressway to Zhuxinzhuang West Road) has obtained the construction permit. At present, the construction enclosure is under construction, which is expected to be completed before the end of 2026. The total length of the road is about 1.7 kilometers, and it is built according to the standards of urban secondary roads.

　　In addition, a number of road projects have been completed or will be started soon. Among them, Anju Road (Qiliqu North Street to Beiqing Road) has been completed, with a total length of about 1.2 kilometers. Qixin North Street (Zhuxinzhuang West Road to Qiyan Road) is about 1.67 kilometers long and has entered the stage of completion acceptance. Zhuxinzhuang Middle Road (Qixin North Street to Dingsi Road) is planned as an urban secondary trunk road, which is scheduled to start in the fourth quarter of 2025 and be completed before the end of the year.

　　Zhuxinzhuang area is located in the west of the future science city. After the completion of the above road, it will effectively connect Zhuxinzhuang Science and Technology Innovation Core and Science and Technology Service Center in series, injecting new kinetic energy into innovation and development. (Zhao Yuhan)

　　Cctv newsAccording to China Consumers Association, in 2023, "double 11" entered its 15th year. At present, the consumer market continues to pick up, and consumer demand tends to be diversified and personalized. In this year’s "double 11", competition for "price power" is the main theme of business marketing, and the pursuit of "cost performance" has become a new trend of residents’ consumption.

　　According to the monitoring data of the State Post Bureau, from November 1 to 11, 2023, the national postal express delivery enterprises collected 5.264 billion express parcels, up 23.22% year-on-year, and the average daily business volume was 1.4 times that of the normal business. Among them, on November 11th, a total of 639 million express parcels were collected, which was 1.87 times of the business volume on weekdays and increased by 15.76% year-on-year.

　　The promotion methods of e-commerce platform are constantly innovating, paying more attention to simplifying rules and strengthening services. Fifteen years later, conventional routines such as full reduction and 10 billion subsidies have been difficult to arouse consumers’ desire to buy, and platforms and businesses have invariably chosen simpler and more direct "lowest price" shopping. The battle report shows that during the "double 11" period, the number of users, orders and total transactions of Tmall and JD.COM all showed an increasing trend.

　　The China Consumers Association used the Internet public opinion monitoring system to analyze the situation of consumer rights protection from October 20th to November 16th. The data shows that consumer feedback tends to be positive as a whole, and live broadcast and low price have become hot topics for consumers in double 11. Besides the old problems of preferential price dispute, counterfeiting and poor after-sales service, price monopoly and software advertising have also brought consumers a bad consumption experience.

　　Centralized areas of consumer rights protection issues

　　The monitoring found that during the "double 11" in 2023, the problems of consumer rights protection focused on the chaos of live broadcast, the quality of goods, the bad experience of mobile phone software advertising, and the dispute of promotion price.

　　(1) Live broadcast of chaos with goods

　　During the monitoring period, there were 1,565,203 pieces of negative information about "live broadcast with goods", accounting for 47.99% of the spit information, with an average daily information of 55,900 pieces. The amount of information was concentrated in the first half of "double 11", reaching a peak of 94,694 items on October 25th. Price monopoly, vulgar goods, false propaganda and other issues are more prominent.

　　(B) the quality of goods

　　During the monitoring period, there were 870,131 negative information about "commodity quality", accounting for 26.68% of the spit information, and the average daily information was 31,076. The amount of information was concentrated in the early days of "double 11" and reached its peak on October 25th, with 55,629 items. Among them, the problem of counterfeiting in the live broadcast room is outstanding.

　　(C) the bad experience of mobile phone software advertising

　　During the monitoring period, there were 75,663 pieces of negative information about "advertising experience", accounting for 2.32% of the spit information, and the daily average amount of information was 2,702 pieces. With the passage of time, the number of Internet users’ spits increased gradually, reaching a peak of 8,552 on November 14th. People’s "slot point" focuses on many forced jump ways of mobile phone software advertisements, which makes consumers hard to prevent and has a particularly headache.

　　(D) Promotion price disputes

　　During the monitoring period, there were 63,757 pieces of negative information about "promotional price disputes", accounting for 1.96% of the spit information, with an average of 2,277 pieces per day. The overall trend fluctuated and rose. On October 31, November 10 and November 13, spit was concentrated, with the highest amount of information on November 10, at 4,635 pieces. The topics are mainly about changing preferential policies at will, suddenly reducing prices near the end of the activity, not insuring prices, inducing consumers to cancel orders and so on.

　　Suggestions on consumer rights protection and public opinion

　　First, platforms and businesses should aim at "price power" and continue to exert their strength. Adhere to legal compliance, tap consumer demand, and provide consumers with more cost-effective products; At the same time, we seek to improve the "service power" and win the favor of consumers with better service and low price strategy.

　　Second, the platform should earnestly fulfill its management responsibilities and maintain market order. On the one hand, give more support to small and medium-sized businesses, reduce service fees, provide support policies and conduct professional training to stimulate market vitality. On the other hand, we should strictly supervise false propaganda and price misleading, eliminate the chaos of live broadcast, improve the professional quality of anchors and create a level playing field.

　　Third, the regulatory authorities should continue to improve their regulatory capacity and level in the field of live delivery. Make good use of scientific and technological strength, use emerging technologies to develop new means of supervision, and further strengthen the standardized management of platforms, anchors, businesses, etc. Be wary of "price monopoly" and "new routines" to ensure consumers’ peace of mind.

　　Fourth, the media should play a greater role in guiding consumer behavior and helping consumer rights protection. Guide consumers to establish a correct consumption concept through public opinion propaganda, improve consumers’ awareness of discrimination, and enhance their resistance to induced consumption; Constantly broaden the channels of consumer rights protection, pay attention to the information of consumer rights protection on social media platforms, guide and help consumers to rationally protect their rights, and provide consumers with more rights protection.

On December 6th, the hot list of short plays was officially announced, and the popular drama "I am a drug god" successfully reached the top and became the focus of audience’s hot discussion. This short play not only attracts people’s attention because of its profound theme, but also gives the broad audience a new perspective on life thinking. With the continuous development of the short play market, the audience’s demand changes gradually, and the choice and preference for content are also evolving. In this context, we might as well discuss the reasons why I am a drug god can achieve such excellent results and the overall development trend of the short drama industry.

"I’m a Drug God" quickly captured the hearts of the audience by deeply discussing the complex relationship between family, affection and human nature. This drama not only combines medical problems with philosophy of life, but also shows people’s struggle and reflection in desperation, and also strengthens their emotional resonance with a real story structure. Many viewers said that after watching the drama, they had mixed feelings, and many people shared their feelings on social media, which further promoted the discussion of the drama. In the hot list of short plays, the success of "I am a drug god" is not only a contest between script and acting, but also a keen capture of social hot topics, which makes the audience feel resonance in reflection.

Followed by another short play "Filial Piety and Father’s Safety", it won the favor of many audiences with warm family stories and plots close to life. By telling the mutual understanding and support between family members, the play makes people re-examine the value of family affection in their busy lives. Many viewers have a strong sense of substitution when watching, and they all say that this drama is like a mirror, reflecting their own life and family relationship. This warm atmosphere has undoubtedly accumulated a lot of positive reputation for this short play, making it rank second in the list.

The third-ranked book, "I am still a stranger", has touched many audiences with its unique narrative style and profound emotional expression. The narrative technique of the play is novel, and a large part of the time is used to show the inner struggle and growth of the characters, giving the audience more room for thinking. It is not only a simple love story, but also deeply discusses the relationship between people and the diversity of emotions. This novel idea is not only refreshing, but also provides a new creative direction for the development of short plays.

With the cooperation of New Wrist and Whip Niu Shi Technology Media, we launched the WETRUE short play popularity list, aiming at providing more powerful market information and direction guidance for short play creators. The daily update of this list ensures that the industry can get fresh market trends and changes in audience tastes in time. Such digital support is very important for the development of short drama industry, which not only helps creators understand the needs of the audience, but also provides impetus for the prosperity and progress of the whole short drama market.

What is even more exciting is that the second short play conference will be held soon by New Wrist. This grand event will not only provide a platform for outstanding short drama creators to display their talents, but also promote exchanges and cooperation in the short drama industry. During the conference, creators and filmmakers can share their creative experiences and insights, and discuss future development trends, thus promoting the continuous progress of the industry. On the one hand, such a platform gives more excellent works a chance to be displayed; On the other hand, it can also inspire more new generation creators to try and inject fresh blood into the short drama industry.

The rapid development of the short play industry means a huge expansion of creative space. With the diversified needs of the audience, short play works must constantly break through in emotional depth and narrative innovation. Short plays, as a new form of artistic expression, are no longer just cultural products that are rapidly consumed, but carriers that bear the creators’ profound thoughts on life and society. This trend has been fully reflected in works such as "I am really a drug god", which also points out the direction for future creation.

In this highly competitive market, the voice of the audience is always the most important weather vane. With the continuous enrichment of short drama categories, the audience’s pursuit of quality and emotion will become more and more obvious. Stories that can directly hit people’s hearts are certainly the direction that short drama creators need to explore constantly. The success of "I am a drug god" also reminds us that excellent works often come from in-depth social thinking and real humanistic care, which are undoubtedly an important driving force for the sustained development of the short drama market.

In the future, with the sustained development of the short drama industry, we have reason to believe that more outstanding works will emerge continuously, bringing richer and more touching audio-visual feasts to the audience. This process is not only a challenge to content creation, but also a promotion to the whole industry. I believe that every creator can find his own position in the feast of this short play and create touching works.

Editor in charge:

Editor’s note: "Hundred years is the first." Speaking of Chinese New Year, what we think is grand and warm, and what we think is rich and colorful. Different nationalities and regions have different years; Different historical periods have passed different years. On the occasion of the Spring Festival, with the theme of "We celebrate the New Year like this", we hope to present readers with a picture of the New Year in which culture, history and personal feelings are intertwined by telling the past of the New Year in a specific time and space.

Yang Zheng (son of Yang Fuxing, the last chieftain of Zhuoni, served as the county magistrate of zhuoni from 1986 to 1989).

I was born in a toast family. My ancestors came from Tibet, and they are said to be descendants of Songzan Gambu. In the Yongle period of the Ming Dynasty, my ancestors joined the Ming Dynasty and were named Zhuoni Tusi. When Emperor Zhengde arrived, they gave our family a surname of Yang. From the founding of New China in 1949 to my father, Yang Fuxing, it has been passed down for 20 generations and experienced the Ming, Qing and Republic of China. In September 1949, my father led an electrifying uprising, which declared the end of the chieftain system.

When I was a child, it was a very happy thing to celebrate the New Year. I remember that there were delicious foods every day from the 23rd candy. Every day, there were melon seeds, walnuts, peanuts, dried persimmons, soybeans (called Tiefabean in Beijing), hot pot and jiaozi for dinner. At that time, there were many pheasants, and they also ate fried pheasants. The pheasant was called by my father himself. On the first day of the Lunar New Year, I will go to the Zhuoni Zen Temple to worship Buddha at about 3 o’clock. I followed it once when I was a child. I only remember that the Dajingtang was dark, which was a bit scary. When we come back, it will be dawn the day after tomorrow. Everyone will shoot guns and adults will pay New Year greetings to each other. We children will remember to collect more lucky money. On the second day of junior high school, my father took us riding to pay a New Year call to my grandmothers. My grandfather had four wives, and three of them lived in different places after the founding of New China. I like it better because I can ride horses and walk a long way. Then, on the eighth day of the first month, we will go to worship the ancestral graves, kowtow, burn incense, burn paper and shoot guns. At that time, guns were not strictly controlled, and they were usually fired into the air, with submachine guns and 79 rifles. I dare not let go of these guns. The recoil is too strong, so the adults hold them and let me pull the trigger. There is a kind of carbine, which has a small recoil. I can hold it myself.

On the fifteenth day of the first month, there will be a sun-bathing activity in Zhuoni. Groups of lamas, surrounded by many people, carry a long scroll of Buddha statues about ten meters, and then slowly put it down. Under the umbrella of the yellow luogai, a living Buddha will lead, surrounded by some lamas chanting, and the outer layer is the masses. My father and I were arranged next to the living Buddha, and there was a carpet on which we sat.

On the 16th day of the first month, it was a "war pole", which was held in the square in front of the Great Sutra Hall of Zen Temple. This activity is held every three years according to religious regulations. I participated in it once in the 1950s, and my father and I were arranged in the cabinet room above the Great Classics Hall. At first, it was a religious dance. The lamas wore masks, colorful silk clothes and some instruments in their hands, and danced with the simple drums and the percussion of cymbals. There should be three or four kinds of French dances, which are performed in turn. An hour or two later, the monks with the most morality and prestige were selected by the temple to fight the pole. At first, a large oil pan was set up in the courtyard, which was filled with green oil, and dry wood was set up under it to burn. On top of the oil pan, a long rod was set up, and a piece of paper with spells was hung on the tip of the rod. The monk murmured, holding an iron spoon with inflammable and explosive substances such as sulfur and salt in it. When the oil in the oil pan was boiled by the fire, the monk suddenly poured the flammable substances in the spoon into the oil pan. Suddenly, the smoke in the pan exploded and the spell on the long pole flew away. At this time, the crowd cheered everywhere, and the people with guns fired their guns into the air, and the activity reached its climax. A monk who pours flammable materials into a frying pan must be profound, or he will hurt himself. It is said that this activity has not been held since 1958, and it was not until the Spring Festival in 1986 that the Zen Temple was held again. At that time, I had graduated from college, first as a teacher in County No.1 Middle School, and then as a deputy magistrate at the end of 1983. During the Spring Festival in 1986, I was the deputy magistrate and had another Spring Festival in Zhuoni. A social fire performance was held during the Spring Festival.Three social fire teams were organized by three villages around the county seat to perform dragon dance, lion dance and stilt walking in the courtyards and streets of various organs. On the 16th day of the first month, the Zen Temple also resumed the activities of fighting poles that had been interrupted for many years. This time, the cabinet room above the Grand Classics Hall was not opened. Several leaders in our county set up a row of tables at the entrance of the Grand Classics Hall to watch the activities. At the end of the activities, no guns were fired, but firecrackers were set off instead.

It is worth mentioning that in 1969, I jumped the queue to Yan ‘an, in the Lanta Brigade of Liangcun Commune in Yan ‘an County. At that time, I advocated a "revolutionary" Spring Festival. On the morning of the first day of junior high school, more than a dozen educated youths took up the burden to send manure to the field. Before noon, the captain announced that we should call it a day, and informed the educated youths to take turns to eat at 30 villagers’ homes in the village during the Spring Festival. Every family entertained the educated youth with the best things. Some families borrowed food from relatives and neighbors when they were out of food, which made us deeply feel the kindness, simplicity and enthusiasm of Yan ‘an people. I’m 65 years old, and I’ve eaten a lot of sumptuous meals all my life, but I always remember the oil cake, sauerkraut vermicelli and millet dry rice I ate in Lanta Brigade during the New Year.

Originally published in the second issue of Vertical and Horizontal in 2016.

Produced by Vertical and Horizontal magazine under China Literature and History Publishing House.

Editor/Yu Yang

For more information, please contact our hotline: 010-81136601 010-81136698 010-81136697.　

Thank you for paying attention to our official micro: China Literature and History Publishing House (micro-signal: wspress1980)

Reporter Wu Linhua

The reporter learned from the press conference held by the Municipal Public Security Bureau yesterday that in 2023, the number of traffic order 110 alarms in Shencheng decreased steadily month by month, making it the only provincial administrative region in China that has not suffered a traffic accident with three or more deaths for three consecutive years. According to a report from a third-party organization, Shanghai’s commuter congestion index decreased by 6.64% last year compared with the same period of last year.

In 2023, the police and the Municipal Transportation Commission will comprehensively manage more than 130 traffic congestion points in the city, scientifically optimize traffic signals at more than 700 intersections, and effectively improve road traffic efficiency. For example, there is a large traffic volume near the ramp on the Shanghai-Min elevated Longchao Road, and there is a conflict between the up-and-down ramps. After detailed investigation and animation simulation, the police adjusted the long solid line from the Shanghai-Min elevated to the inner ring, increased the parking space on the ramp, and optimized the lane warning signs to make drivers more clear about the lane guidance, and at the same time increased the police force to strengthen the "up-and-down linkage" management during peak hours. After the implementation of the measures, the average traffic flow on the east main line of Humin Elevated Road (Liuzhou Road to Caobao Road) and the east ramp of Longcao Road increased by about 10% compared with that before the reconstruction, and the average speed during the peak hours also increased by 15%.

Shanghai police also promoted traffic safety hidden danger management and enterprise traffic safety supervision in a rolling way around the five dimensions of "people, vehicles, roads, enterprises and time periods", achieving a "double decline" in the number of deaths and injuries in road traffic accidents throughout the year. The city has completed more than 1,000 high-risk roads facing water hazards, installed more than 900 sets of "intersection sentries", and promoted more than 26,000 large trucks to install monitoring and early warning equipment for right-turn blind spots. The installation rate of warning devices for long-distance buses and tourist buses passengers without seat belts reached 100%.

In order to strengthen the source management of road safety hazards, since 2023, the police have pushed key transportation enterprises to face-to-face record one by one, updated the information of enterprises, vehicles and drivers in real time, assessed and monitored the risk level of enterprises based on relevant data models, and urged more than 4,000 enterprises with hidden problems to rectify in time.

In addition, the police conducted strict in-depth investigations on 51 road traffic accidents involving enterprises, and took criminal compulsory measures against four responsible persons of enterprises and imposed administrative fines on 37 enterprises for the problems exposed behind illegal acts such as overloading of transport vehicles, non-stop turning right and occupation of non-motor vehicle lanes, and safety education and training became a mere formality, incomplete account records or falsification.

Recently, in order to maintain the road traffic order at the end of the year and the beginning of the year, the police relied on big data analysis and judgment to carry out accurate investigation and rectification actions against traffic violations, focusing on large vehicles that have not been inspected annually, vehicles that are suspected of fatigue driving "two passengers, one danger and one cargo" and trucks that have not been scrapped after the deadline, but have been cancelled but still driving on the road, achieving a precise blow to "hidden violations". At present, more than 5,800 "hidden violations" have been investigated.

According to the data analysis, the police also sorted out 69 key stations and enterprises of large freight cars, as well as 170 illegal high-risk points of large freight cars, and strengthened on-site law enforcement inspections for targeted police. The police also cooperated with traffic law enforcement departments to carry out joint rectification, and actively investigated and dealt with traffic violations that may endanger safety, such as fatigue driving of network car drivers.

Shanghai police said that the new year will rely on the help of science and technology, innovate governance methods, and start from the nuances to create a high-quality traffic environment for high-quality economic development. In 2024, the city will complete the renovation of no less than 100 road congestion points, promote the "low-light composite electronic police" and "no-light electronic police", and strive to eliminate the influence of fill light on drivers; It will further deepen the "streamline administration, delegate power, strengthen regulation and improve services" reform, and promote the decentralization of car driving business to traffic police detachments, "comprehensive windows" of police stations and socialized service institutions, so as to facilitate the masses to "do it nearby" and "do it at once".

Ada’s hair is dazzling.

Ada’s eyes are charming

Ada is sexy and mysterious.

Ada career line is half covered and half exposed.

Ada is lazy.

Ada has a sweet smile.

1905 movie network news On May 15th, Ada released a set of black and white beautiful pictures in Weibo. A low-necked sleeveless vest sets off Ada’s graceful figure, and her dazzling hair makes people feel lazy and full of mystery and sexy. Her charming eyes are mature and elegant, and her career line is half covered and half exposed, which has a hazy beauty of "still hiding half her face from us behind her guitar".

Netizens have commented that "Ada is a sexy goddess and won’t accept refutation", "I can’t describe my sister’s beauty, so I can only praise Brother Bin’s photography skills" and "It feels like a Hollywood star photo in the 1960s, which is so beautiful".

Wisdom things (WeChat official account: zhidxcom)
Wen | Xinyuan

In 2019, the new cloud AI chip battlefield is surging.

In the past few years, artificial intelligence (AI) has suddenly exploded from a neglected academic unpopular research to the forefront of commercialization in running all the way, which has set off a hurricane of intelligent upgrading and Internet of Everything in various fields closely related to people’s lives, such as security, finance, education, manufacturing, home and entertainment.

The direct promoters of this unprecedented technological revolution are foreign Internet giants such as Google, Microsoft, Facebook and domestic BAT, as well as a number of new AI start-ups, and the soul pillar of these companies to quickly break ground in the AI field is to provide a steady stream of AI hardware providers with high-density computing power.

AI hardware application scenarios are usually divided into cloud and terminal. Cloud mainly refers to large-scale data centers and servers, and terminals include rich scenarios such as mobile phones, vehicles, security cameras and robots.

Whether it is online translation, voice assistant, personalized recommendation or various AI development platforms that lower the threshold for developers to use, wherever AI technology is needed, cloud AI chips are needed to provide strong computing support for data centers around the clock.

According to the data revealed by NVIDIA in 2017, by 2020, the global market size of cloud AI chips will exceed 20 billion US dollars, and this huge market has become a place where chip giants are eyeing.

NVIDIA General Graphics Processing Unit (GPGPU) is soaring on the east wind of deep learning. Its share price was still $20 in 2015, and soared to $292 in October 2018. Its market value surpassed KFC and McDonald’s, and it became the first share in the AI field, with a market value of billions of dollars, enjoying unlimited scenery.

Its rocket-like rise awakened a number of potential competitors, and the storm appeared on the horizon. Semiconductor giants such as Intel and AMD are catching up. Google, Amazon, Baidu, and Huawei have made cross-border self-research, and dozens of new chip startups have sprung up, with the intention of breaking through the ceiling of cloud AI chip performance and reshaping this market through self-research architecture.

This paper will take a panoramic view of the battle of cloud AI chips, take stock of the five semiconductor giants, seven Chinese and American technology giants and 20 domestic and foreign chip companies that have joined the battle, and see if NVIDIA, which once created myths, can maintain its legendary empire? Can the new computing architecture that has appeared or is being developed now adapt to the future algorithms? Which companies are more likely to survive in a competitive environment with strong hands?

Whoever can dominate this cloud AI chip war will win more say in the battle of cloud computing and AI market in the future.

It all started with an accident, and it was by no means an accident.

More than a decade ago, NVIDIA and AMD became the two dominant players in the field of graphics cards after fierce fighting with dozens of rivals. At that time, most NVIDIA employees did not know what artificial intelligence (AI) was.

At that time, NVIDIA’s total revenue was about $3 billion, and its founder and CEO Huang Renxun made a risky decision-spending $500 million on CUDA projects every year, and transforming GPU into a more general computing tool through a series of changes and software development, with a total amount of nearly $10 billion.

This is a very forward-looking decision. In 2006, CUDA, the world’s first general computing solution on GPU, came into existence. This technology brought more and more convenient entry experience to programmers, and gradually accumulated a strong and stable developer ecology for NVIDIA GPU.

Until 2012, NVIDIA encountered the enthusiasm of deep learning.

This year, Geoffrey Hinton, a professor at the University of Toronto, Canada, a master in the field of machine learning and the father of neural networks, led the research group to train CNN)AlexNet, a convolutional neural network (CNN), and won the ImageNet image recognition competition in one fell swoop, pushing AI to the historical turning point of academic focus.

GPU is not born for deep learning, and its parallel computing ability coincides with the logic of deep learning algorithm. Each GPU has thousands of cores in parallel, and these cores usually perform many low-level and complicated mathematical operations, which is very suitable for running deep learning algorithms.

After that, the increasingly strong "CUDA+GPU" combination, with its invincible processing speed and multitasking ability, quickly captured the hearts of a large number of researchers and soon became an essential component of major data centers and cloud service infrastructures around the world.

The battle of the giants’ cloud AI chips quietly kicked off.

With an early start and ecological stability, NVIDIA soon became the leader in the cloud AI chip market.

NVIDIA is marching forward on the road to stronger, showing amazing technologies such as Tensor Core and NVSwitch one after another, and constantly creating new performance benchmarks. In addition, it also builds a GPU cloud, which enables developers to download a new version of the deep learning optimization software stack container at any time, greatly lowering the threshold for AI R&D and application.

In this way, NVIDIA built an indestructible wall by accumulating time, talents and technology. Anyone who wants a city needs to follow the rules specified by NVIDIA. As of today, NVIDIA has more than 10,000 engineers, and its GPU+CUDA computing platform is by far the most mature AI training program, which devours the cakes in most training markets.

From the functional point of view, the cloud AI chip is mainly doing two things: Training and Inference.

Training is to cram massive data into the machine and make it learn to master specific functions by repeatedly adjusting the AI algorithm. This process requires extremely high computational performance, accuracy and universality.

Reasoning is to apply the trained model, its parameters have been solidified, and it does not need massive data, so the requirements for performance, accuracy and universality are not as high as those of training.

GPU is a difficult mountain to climb in the training market, but its advantages are relatively less obvious in the reasoning market which requires higher power consumption.

And here, it is also the direction where the semiconductor giants who entered the game late gathered.

▲ Incomplete statistics of major cloud AI chip products of chip giants

Chips are a winner-take-all market, and cloud AI chips are no exception. NVIDIA’s high, medium and low-end general-purpose GPU for accelerating data center applications has always been a performance benchmark for all players.

NVIDIA invested billions of dollars and thousands of engineers in a short time, and launched the first Pascal GPU optimized for deep learning in 2016. In 2017, it launched Volta, a new GPU architecture with five times higher performance than Pascal, and TensorRT 3, a neural network reasoning accelerator, also appeared at the same time.

In the latest quarterly financial report, NVIDIA’s data center revenue increased by 58% year-on-year to $792 million, accounting for nearly 25% of the company’s total revenue, reaching a total of $2.86 billion in the past four quarters. If it can maintain this growth, it is estimated that the data center will reach about $4.5 billion in 2019.

AMD, which has long competed with NVIDIA in the GPU field, is also actively promoting the research and development of AI accelerated computing. In December 2016, AMD announced ——Radeon Instinct, an accelerator card program focusing on AI and deep learning.

Speaking of it, AMD’s start in the field of deep learning is inseparable from the support of China. Baidu was the first China company to adopt AMD Radeon Instinct GPU in its data center, and later Alibaba signed a contract with AMD.

At present, AMD’s GPU is still at least behind NVIDIA’s Tesla V100. However, when NVIDIA’s new move was not issued, AMD took the lead in announcing the launch of the world’s first 7nm GPU at its Next Horizon conference, named Radeon Instinct MI60, with a memory bandwidth of 1 TB/s. It also claims that its 7nm GPU has become the fastest double-precision accelerator in the world through AMD Infinity Fabric Link and other technologies, which can provide floating-point performance as high as 7.4 TFLOPS.

In addition to providing GPU chips, AMD is also building a more powerful open source machine learning ecosystem by launching ROCm open software platform.

Although GPU can’t resist NVIDIA for the time being, AMD has its own unique advantages. AMD has both GPU and CPU, and it can connect seamlessly between GPU and CPU with Infinity Fabric, but it is difficult for Intel Xeon processor +NVIDIA GPU to achieve such a perfect connection.

Imagination Technologies is also camped in GPU market, but it has been deeply involved in mobile GPU for a long time. From 2017 to 2018, Imagination announced three new PowerVR graphics processing units (GPUs), focusing on the AI terminal market.

At the end of last year, Imagination executives revealed in an interview that Imagination may announce the launch of GPU for AI training.

In the application of AI reasoning, FPGA has the advantages of flexibility and programmability compared with ASIC, which can be reconfigured in real time for specific tasks and has lower power consumption than GPU.

▲ Flexibility and performance difference of processors

The eldest and second children in the field of FPGA are Xilinx and Intel Altera all the year round. Facing the emerging AI market, the innovative genes in the body are also eager to try.

Xilinx’s killer is called Versal, which is the first adaptive computing acceleration platform (ACAP) in the industry. It adopts TSMC’s 7nm process and integrates AI and DSP engines. Its software and hardware can be programmed and optimized by developers.

This killer took four years to polish, and it is said that the AI inference performance of Versal AI Core is expected to be 8 times higher than that of the industry-leading GPU. According to the news released by Xilinx, Versal will deliver the goods this year.

Some insiders believe that Versal series may change the AI reasoning market.

If NVIDIA opens the door to AI by natural genes, then Intel quickly ranks among the front rows of cloud AI chips by the shortcut of "buy in buy buy". As a semiconductor overlord for decades, Intel’s first goal is to become a "generalist".

As we all know, Intel’s invincible trump card is Xeon processor. Xeon processor is like a brilliant strategist, strategizing and able to handle all kinds of tasks, but if you let him forge weapons, his efficiency is completely inferior to that of a warrior with a simple mind but a brute force.

Therefore, in the face of AI with a large number of repetitive and simple operations, it is both overqualified and inefficient for Xeon processors to handle such tasks. Intel’s approach is to match the Xeon processor with an accelerator.

What if I don’t have the technical background to be an AI accelerator? With a stroke of Intel’s pen, buy it directly!

In December 2015, Intel spent $16.7 billion to buy Altera, the second child of the programmable logic device (FPGA) at that time. Now Intel has accelerated some tasks in the data center by more than ten times with the "Xeon+Altera FPGA" heterogeneous chip.

Especially in the past year, Intel’s overweight on FPGA is visible to the naked eye. Two years ago, Intel successively launched Stratix 10 series, which is known as the fastest FPGA chip in history, and this series won the favor of Microsoft.

Microsoft launched Project Brainwave, a cloud solution based on Intel Stratix 10 FPGA, claiming that it runs at a speed of 39.5 TFLOPS with a delay of less than 1 ms.

In addition to Stratix 10 FPGA chip, Intel first settled in Chongqing in December last year, and then in April this year, it unveiled a new weapon that has been quietly polished for several years-FPGA Agilex with a new architecture, which integrated Intel’s most advanced 10nm process, 3D packaging, the second generation HyperFlex and other innovative technologies.

Intel’s FPGA has gained a foothold in the server market, while another important transaction is still in the dormant period.

In August 2016, Intel spent 300-400 million dollars to buy Nervana, a California startup dedicated to building deep learning hardware. Shortly after the acquisition, the former CEO of Nervana was promoted to the general manager of Intel AI business unit, and the first deep learning special chip Lake Crest using TSMC’s 28nm process was mass-produced in 2018, and claimed that its performance was 10 times that of the fastest GPU at that time.

In May 2018, Intel’s new cloud AI chip Nervana Neural Network Processors (NNP)-Spring Crest was officially unveiled. It is said that its power consumption is less than 210 watts, and its training performance is 3-4 times higher than that of Lake Crest, which will be open to users in the second half of 2019.

For cloud AI chip reasoning, Intel revealed at CES in Las Vegas that it is working closely with Facebook on the reasoning version of Nervana neural network processor NNP-I. NNP-I will be a system-on-a-chip (SoC) with built-in Intel 10nm transistors, and will include the IceLake x86 core.

Compared with Google’s TPU, Carey Kloss, vice president of Intel Artificial Intelligence Group (AIPG) and core member of Nervana team, thinks that TPU 2.0 is similar to Lake Crest and TPU 3.0 is similar to Spring Crest.

Qualcomm, which is flourishing in the field of mobile chips, has just raised a stepping stone to enter the field of cloud computing and supercomputing.

In April this year, Qualcomm announced the launch of the Cloud AI 100 accelerator, which will expand Qualcomm’s technology to data centers, and it is expected that samples will be delivered to customers in the second half of 2019.

It is reported that this accelerator is based on Qualcomm’s technology accumulation in signal processing and efficacy, and is specially designed to meet the rapidly increasing demand of AI reasoning processing in the cloud, so that distributed intelligence can spread from the cloud to the user’s edge terminal and all nodes between the cloud and the edge terminal.

Keith Kressin, senior vice president of product management in Qualcomm, said: "Qualcomm CloudAI 100 accelerator will set a new benchmark for AI inference processors in data centers in today’s industry-no matter which combination of CPU, GPU and/or FPGA is adopted to realize AI inference processors."

In addition, he also said that Qualcomm is currently in an advantageous position to support complete AI solutions from the cloud to the edge, and all AI solutions can be connected with 5G with the advantages of high speed and low latency.

Compared with the ambitious chip giants facing the cloud and data center market, the minds of the following cross-border players are relatively "simple".

The goal of these Chinese and American Internet giants is not to directly compete with NVIDIA, Intel or AMD, but to provide their own cloud customers with powerful computing power and reduce their dependence on traditional chip manufacturers.

Their choice of self-developed chips is also different. Google, Amazon and others choose the route of ASIC, while Microsoft and others are committed to using field programmable gate array (FPGA).

▲ Incomplete statistics of major cloud AI chip products of cross-border technology giants

As one of the earliest technology companies to start AI-related research and development, Google is also a pioneer in testing special AI chips, and it was the first to verify that ASIC can replace GPU in the field of deep learning.

In 2016, Google launched its own AI chip Tensor Processing Unit(TPU), which has now entered the third generation, providing computing support for various AI applications such as Google’s voice assistant, Google Maps and Google Translation. The TPU originally designed is used in the reasoning stage of deep learning, and the new version can already be used for AI training.

Google claims that it takes one day to train a machine translation system using 32 best commercial GPUs, and the same workload takes six hours on eight connected TPU.

Google currently only operates this device in its own data center and does not sell it to the outside world. Recently, however, Google said it would allow other companies to purchase its TPU chips through its cloud computing service.

Google TPU is limited in Google’s external service market. TPU can only be used and run with Google TensorFlow AI framework. Users cannot use them to train or run AI built with Apache MxNet or Facebook’s PyTorch, nor can they be used in non-AI HPC applications where GPU occupies the supreme position.

But Google is satisfied with this because it regards TPU and TensorFlow as its comprehensive AI leadership strategy. Software optimized for its software is optimized for its software, which can build a powerful and durable platform.

The new news this year is that Google set up a new chip team gChips in Bangalore, and recruited at least 16 technical veterans from traditional chip companies such as Intel, Qualcomm, Broadcom and NVIDIA.

In May last year, the Microsoft AI chip Brainwave opened the cloud beta, saying that the FPGA chip used in the Project Brainwave computing platform was designed for real-time AI, which was five times faster than the TPU chip used by Google (the Microsoft AI chip Brainwave open cloud beta was five times faster than TPU). Jason Zander, executive vice president of Microsoft Azure, also said that Microsoft Azure actually designed many self-developed chips for data centers.

I have to admit that the domestic technology giants named the chip, and the cultural Level was higher than that of foreign countries by more than one level.

The "Kunlun" named by Baidu for the cloud AI chip is the first mountain in China. According to legend, the ancestor of this mountain was honored as "the ancestor of thousands of mountains" and "the ancestor of Long Mai" by the ancients, and well-known myths and legends such as the Goddess Chang’e flying to the moon, Journey to the West and Legend of the White Snake are all related to this mountain.
The "rise" of Huawei’s cloud AI chip takes the meaning of transcending the world, rising and imposing, and is quite popular among literati.

Both Baidu and Huawei are domestic technology companies that have made cross-border cores early. As early as August 2017, Baidu released a 256-core FPGA-based cloud computing acceleration chip at the Hot Chips conference in California, and its partner was Xilinx. Huawei made chips earlier. In 2004, HiSilicon, a semiconductor company, was established, but it used to be a chip solution for terminals.

In the second half of 2018, a new round of core-making forces represented by them blew the horn of China cloud AI chip charge.

Baidu is an early technology giant in China. As early as 2010, it began to use FPGA to do research and development of AI architecture. In 2011, it launched small-scale deployment. In 2015, it broke the deployment scale of thousands of chips. In 2017, it deployed more than 10,000 FPGAs. Baidu’s internal data center and autonomous driving system were all used on a large scale.

In August 2017, Baidu released a 256-core XPU chip based on FPGA, which is in cooperation with Xilinx. The core is small, there is no cache or operating system, and the efficiency is equivalent to that of CPU.

Then at the Baidu AI Developers Conference held in July 2018, Baidu announced Kunlun, the most powerful AI chip in the industry at that time.

In terms of parameters, Kunlun chips are manufactured by Samsung, using 14nm technology, with a memory bandwidth of 512GB/s and tens of thousands of cores, which can provide computing power of 260 TOPS at a power consumption of over 100W W..

Compared with NVIDIA’s latest Turing architecture T4 GPU, T4′ s maximum power consumption is 70W, and the highest computing power it can provide is 260 TOPS. However, this GPU was released two months later than Kunlun Chip, and it was not sold in China at the beginning. Ouyang Jian, chief architect of Baidu, revealed at this year’s AI chip innovation summit that this year’s "Kunlun" will be used on a large scale within Baidu.

Huawei’s cloud AI chip Ascension 910 is directly at the release site and the front PK of NVIDIA and Google. Ascent 910 directly uses the most advanced 7nm technology, adopts the Leonardo da Vinci architecture developed by Huawei, and the maximum power consumption is 350W. Huawei’s banner is "the chip with the highest computing density on a single chip" as of the release date, and its semi-precision (FP16) computing power reaches 256 TFLOPS, which is twice as high as the 125 TFLOPS of NVIDIA V100.

Xu Zhijun even said that if 1024 Ascending 910s are collected, there will be "the largest AI computing cluster in the world so far, with a performance of 256P, and no matter how complicated the model is, it can be easily trained." This large-scale distributed training system is called "Ascend Cluster".

In terms of landing, Baidu said that its Kunlun will be widely used in Baidu data center this year. Huawei’s Ascent 910 was originally planned to be listed in Q2 this year. Now, under the background of trade war, I don’t know if it will be delayed.

As the leaders of the cloud computing market in China and the United States, Alibaba and Amazon are a little late, but they will never be absent.

The research and development purposes of the two companies are very clear, which are to solve the AI reasoning operation problems of commercial scenes such as image, video recognition and cloud computing, improve the operation efficiency and reduce the cost.

Alibaba Dharma Institute announced in April last year that the performance of Ali-NPU will be 10 times that of mainstream CPU and GPU architecture AI chips on the market, with manufacturing cost and power consumption only half, and cost performance exceeding 40 times. In the same month, Ali wholly acquired Zhongtianwei, the only independent embedded CPU IP core company in mainland China.

The new progress occurred in September, when Ali merged Zhongtianwei and Dharma Institute’s self-developed chip business into a chip company, Pingtou Ge. The important task of developing Ali-NPU is taken over by Pingtou Ge. The first batch of AI chips are expected to be available in the second half of 2019 and will be applied in cloud data scenarios such as Ali data center, urban brain and autonomous driving. It will be opened to the outside world through Alibaba Cloud in the future.

In the simulation verification test, the prototype of this chip saved 35% of the hardware cost of laying Ali city brain. But since then, Ali has hardly made any sound of relevant progress.

Amazon’s cloud AI chip Inferentia was announced at the Re: Invent conference held in Las Vegas last November.

The technical source of this chip can be traced back to Annapurna Labs, an Israeli chip company acquired by Amazon for $350 million in early 2015. According to the official introduction, each Infirentia chip provides hundreds of TOPS, and multiple AWS Inferentia chips can form thousands of TOPS. The chip is still under development. According to the forecast, the chip will be listed at the end of 2019.

Facebook’s core-building plan surfaced very early, but it is the player with the least information exposure.

In addition to buying relatively mature chip companies, recruiting is also a standing choice. Facebook’s core-building plan first appeared in April last year, and official website posted an advertisement for ASIC&FPGA design engineers to form a chip team. Three months later, the US media Bloomberg reported that Facebook poached Shahriar Rabii, Google’s senior engineer, as vice president and chip leader.

Yann LeCun, the chief artificial intelligence scientist of Facebook and the winner of the latest Turing Award, revealed in an interview that its core-building is mainly to meet the needs of real-time video surveillance of the website in the future.

By January of this year, Intel said at the Global Consumer Electronics Show (CES) that it was working with Facebook to develop a new AI chip to speed up reasoning and strive to complete it in the second half of this year.

But so far, the outside world knows nothing about the performance information of Facebook AI chip.

The revival of AI has subverted the stable situation of the whole industry of carry, a top chip company such as Intel, AMD and Qualcomm, and created opportunities for a new group of chip entrepreneurs.

Some startups want to create a new platform from scratch, all the way to the hardware, which is optimized for AI operation. It is hoped that by doing so, it can surpass GPU in speed, power consumption and even the actual size of the chip.

▲ Incomplete statistics of major cloud AI chip products of domestic start-ups.

Let’s talk about the domestic cloud AI chip creation enterprises, the most dazzling of which are Bitland and Zhongke Cambrian.

Bitcontinent is famous as the leader of mining chip industry, but in the past year’s bitcoin ebb tide, Bitcontinent was the first to fall into the whirlpool of public opinion, and the listing plan failed to be realized as scheduled.

This company, established in 2013, started the AI chip business in 2015. Following the launch of the first generation of 28nm cloud AI chip product BM1680 in 2017, it released the second generation BM1682 in the first quarter of 2018, with an iteration time of only 9 months.

According to the core-making plan announced by Bitcontinent last year, the 12nm cloud chip BM1684 should be launched at the end of 2018, and BM1686 will be launched in 2019, probably using the 7nm process, but both chips are late.

Like Bitland, there are also AI chiplet unicorns, such as the Cambrian.

In CAMBRIAN, the neural network processor (NPU) embedded in Kirin 970, Huawei’s first mobile phone AI chip, became a popular fried chicken in AI chip enterprises at home and abroad. After two rounds of financing, the overall valuation was about 2.5 billion US dollars (about 17 billion yuan).

In May 2018, the CAMBRIAN officially released the first generation cloud AI chip MLU100, which is said to provide better performance than NVIDIA V100 with lower power. Iflytek, its customer, once disclosed the test results, saying that the energy consumption efficiency of MLU100 chip in voice intelligent processing is more than five times ahead of the cloud GPU solution of international competitors.

A year later, Siyuan 270, the second-generation cloud AI chip, did not get hot first, and some of its performance was exposed by netizens in Zhihu. The peak performance and power consumption were basically the same as NVIDIA Tesla T4. It is rumored in the industry that the Cambrian may make a breakthrough in the field of low-precision training. The chip will be released in the near future if nothing unexpected happens.

The startups that want to benchmark NVIDIA and Google don’t stop there.

A slightly surprising player is Etu Technology, one of the four little dragons of domestic computer vision (CV). In May of this year, Yitu released questcore, the first cloud AI chip jointly developed with ThinkForce, an AI chip maker.

Yizhi Electronics is a low-key but not to be underestimated AI chip startup in Shanghai. In 2017, it received a series A financing of 450 million yuan from Eto Technology, Yunfeng Fund, Sequoia Capital and Gaochun Capital. Its core members come from semiconductor giants such as IBM, AMD, Intel, Broadcom, Cadence, etc., and all have more than ten years of experience in the chip industry.

This customized SoC chip for cloud deep learning reasoning adopts 16nm process and ManyCore architecture with independent intellectual property rights. It is said that it can provide visual reasoning performance of up to 15 TOPS per second, and only accelerates for INT 8 data (8-bit integer data type). The maximum power consumption is only 20W, which is smaller than an ordinary light bulb.

According to the figure, the development of this chip is not to pursue the computing power of hundreds of T like NVIDIA, but to value high computing density.

Like the aforementioned cross-border technology giants, the first step of commercialization of Yitu chip is to package and sell it with its own software and hardware and solutions, and it will not be sold separately. The second and third generation products are also in preparation.

Shanghai’s hot new core-making force is Suiyuan Technology. It can be said that it is the youngest AI chip maker in China. It was established in March 2018 and received 340 million yuan of Pre-A financing from Tencent, focusing on R&D investment in cloud AI acceleration chips and related software ecology. This is the first time Tencent has invested in a domestic AI chip venture.

The founding team of Suiyuan Technology mainly came from AMD, and its founder Zhao Lidong previously worked in AMD China, and then went to Rideco (now merged with Spreadtrum to become Ziguang Zhanrui) as the president.

On June 6, 2019, Suiyuan Technology announced a new round of financing of RMB 300 million, which was led by Red Dot Venture Capital China Fund and invested by Haisong Capital and Tencent. The mystery of its deep learning high-end chip has not yet been unveiled.

Different from the previous players, Tianzhixin and Denglin Technology chose a general-purpose GPU that directly matched NVIDIA.

In China, there is no GPGPU company that can compete with NVIDIA, which is an opportunity worth cutting into for entrepreneurs.
The core-making lineups of the two companies are very mature. The hardware team of Tianzhixin is based on AMD’s GPU teams in Shanghai and Silicon Valley, and the founding team of Denglin Technology is also a veteran in the GPU industry for many years.

At present, the high, medium and low-end GPGPU products of Tianzhixin are under development, and its high-end chip Big Island will support cloud reasoning and training at the same time. The GPGPU processor of Denglin Technology has also passed FPGA verification, and the design of the first generation product Goldwasser has been completed, and it is planned to be available for customer testing before the end of this year.

There is also a startup named Longjiazhi, which was founded in July 2017 and led by Zhixin Capital and Yiling Capital, and is committed to the research and development of TPU chips.

In order to meet the requirements of low latency, high reliability and data security, Long Jiazhi introduced a new chip type, Mission-Critical AI Processor. The first generation chip was named Dino-TPU, which was first applied in cloud data centers. Its computing power exceeded all GPUs except the latest Nvidia Volta, with a delay of only 1/10 of that of Volta V100, power consumption of 75W, and unique redundancy backup and data security.

According to Long Jiazhi’s development plan, the company plans to complete the streaming of the first chip by the end of 2018.

On the other side of the ocean, many AI chip startups in the United States have also targeted the cloud and computing center market.

A company with a strong presence last year was Wave Computing. This startup acquired MIPS, an old chip IP supplier, last year, and also launched the MIPS open plan. Its accumulated financing reached $117 million.

Its core product is called Data Stream Processor Unit (DPU), which adopts CGRA (Coarsegrain Reconfigurable Array/Accelerator) technology, and is suitable for large-scale asynchronous parallel computing problems.

Its main advantage is to make hardware more flexible to adapt to software, achieve a good comprehensive balance in programmability (or universality) and performance, lower the threshold of AI chip development, and will not be affected by memory bottlenecks in accelerators such as GPU.

Wave’s first generation DPU adopts 16nm process technology and runs at a speed above 6 GHz, which has been put into commercial use. According CTO Chris Nicol, its senior vice president and CTO, the new generation of 7nm DPU will introduce MIPS technology and adopt HBM(High Band Memory), which is expected to be released next year.

There is also a very mysterious startup, Cerebras System, which was founded in California in 2016. Even if it hasn’t released any products yet, it doesn’t prevent it from being often compared with chip giants.

Cerebras’s founding team mostly comes from chip giant AMD. Andrew Feldman, its co-founder and CEO, previously founded SeaMicro, a low-power server manufacturer, which was acquired by AMD for $334 million in 2012. Since then, Feldman spent two and a half years climbing to the position of vice president of AMD.

Cerebras raised $112 million in three rounds of financing, and its valuation has soared to as high as $860 million. Today, Cerebras is still in secret mode. According to relevant sources, its hardware will be tailored for "training" deep learning algorithms.

▲Cerebras uses deep learning accelerator for neural network training and reasoning patent.

The Groq founding team established in April 2017 is even more eye-catching, with 8 people from the core team of Google TPU. This startup is ambitious as soon as it comes out, and the computing power of official website display chip will reach 400 TOPS.

SambaNova Systems was founded seven months later than Groq, and headquartered in Palo Alto, California. Its founders include two Stanford professors, Kunle Olukotun and ChrisRé, and an old chip company (Sun’s former senior vice president of development).

Its A round of financing was led by Google Venture(GV), the venture capital department of Google’s parent company Alphabet. This is the first time that GV has invested in an artificial intelligence chip company. In April this year, intel capital announced a total of $117 million in new investment in 14 technology startups, and SambaNova Systems was also on the list.

In addition to China and the United States, AI chip startups in other regions are also gaining momentum.

The most optimistic is a well-funded British unicorn Graphcore, which was established in 2016 with a valuation of US$ 1.7 billion and accumulated financing of US$ 312 million. This startup can be called a giant harvester with a strong investment lineup, including Sequoia Capital, BMW, Microsoft, Bosch and Dell Technology.

This company has built an intelligent processing unit (IPU) specially designed for machine intelligent workload, which supports on-chip interconnection and on-chip storage, and extends from edge devices to "Colossus" dual-chip package for data center training and reasoning.
Graphcore wrote in official website: Our IPU system aims to reduce the cost of accelerating AI applications in cloud and enterprise data centers, and improve the performance of training and reasoning by as much as 100 times compared with the fastest system at present.

At the event of NeurIPS at the end of last year, Graphcore showed an example configuration RackScale IPU-Pod, including 32 1U IPU-Machines, each consisting of four Colossus GC2 IPU processors, providing mixed precision calculation of 500 TFLOPS, more than 1.2GB of processor memory and more than 200TB/s memory bandwidth.

▲Graphcore IPU-Pod racksale system

Habana Labs, another Israeli startup founded in 2016, announced at the AI Hardware Summit last September that it was ready to launch its first AI chip Goya for reasoning. It showed the throughput of classifying 15,000 images per second in the Resnet50 image classification database, which was about 50% higher than NVIDIA’s T4 device, with a delay time of 1.3ms and a power consumption of only 100 W.

Its latest $75 million Series B financing (December 2018) was led by Intel Venture Capital, and part of the funds will be used to develop the second chip Gaudi, which will be oriented to the training market. It is said that the training performance can be linearly extended to more than 1,000 processors.

AlphaICs of India was also established in 2016, and is designing AI chips and working on AI 2.0, hoping to realize the next generation of AI through this series of products.

One of the co-founders of AlphaICs is Vinod Dham, who has the title of "Father of Pentium Chip". He cooperated with some young chip designers to create an executable agent-based AI co-processing chip-RAP chip.

Dham said that AlphaICs chips have an advantage over competitors in processing speed, and that most of what we see now belongs to weak AI, and they can be called "strong AI".

According to Dham, the RAP chip is expected to be launched in mid-2019, "hoping to create a big bang for the real AI".

Tenstorrent is a startup located in Toronto, Canada. It was founded by two former AMD engineers, Ljubisa Bajic and Milos Trajkovic. Most of the core teams came from NVIDIA and AMD to develop high-performance processors designed for deep learning and intelligent hardware.

Earlier last year, the company received a seed round investment from Real Ventures, but it is still in a secret mode.

Among the hardware forces in the field of cloud and data center, a special team is favored by domestic and foreign technology giants, which is photonic AI chip.

Different from conventional chips, these chips use photonic circuits instead of electronic transmission signals. They have higher transmission speed, lower delay and higher throughput than electronic circuits.

In 2016, the MIT research team built the first optical computing system, which was published in the top journal Nature Photonics in 2017 as a cover article. It is this paper that inspires more people around the world to invest in the research and development of photonic AI chips.

Only this MIT team hatched two American companies, Lightelligence and LightMatter, in 2017.

In February 2018, Lightelligence received a $10 million seed round financing from Baidu Venture Capital and American semiconductor industry executives. In February 2019, LightMatter received a $22 million B round financing led by Google Ventures, a venture capital department of Google’s parent company Alphabet.

Lightelligence claims that Photonic Circuits can not only be used as a coprocessor of CPU to accelerate deep learning training and reasoning in the field of cloud computing, but also be used for network edge devices that require high efficiency and low energy consumption.

In April this year, Lightelligence announced the successful development of the world’s first photonic chip Prototype board (prototype), and its photonic chips have been in contact with customers at Google, Facebook, AWS and BAT levels.

LightMatter also focuses on large cloud computing data centers and high-performance computing clusters. They have built two early chips, one of which contains more than eleven transistors.

Inspired by the MIT paper, in 2017, the first photonic AI chip enterprise in China was founded by doctoral students from 10 universities including Tsinghua University, Peking University and Beijing Jiaotong University.

The company received angel round financing in September 2018. It is said that the performance of its photonic chip is 1000 times that of electronic chip, and the power consumption is only 1% of that of electronic chip.

Just this month, Bill Gates also began to invest in AI chips, and invested in Luminous, which also developed silicon light technology. Other investors include the 10100 fund of Uber co-founder Travis Kalanick and the current Uber CEO Dara Khosrowshahi.

Luminous currently has only seven members, but its appetite is not small. Its goal is to create a substitute for 3,000 circuit boards containing Google’s latest Tensor Processing Unit AI chip. Their methods draw lessons from the early work of their co-founder Mitchell Nahmias in neuroporphology photonics at Princeton University.

Now the common problem of these startups is that it is not clear how long it will take to release the first mass-produced photonic AI chips, and whether the practical application effect of these chips can really replace the position of electronic chips.

Nowadays, there are dozens of players who have cut into the cloud AI chip market. However, the general pattern of the software, hardware and service market, which is dominated by NVIDIA and divided by many semiconductor giants, is still relatively stable, and it is not an easy task to produce new pattern changes.

For the chip industry, sufficient production capacity is crucial.

Semiconductor giants can achieve 10 times or 100 times the production capacity, but it is difficult for a startup to do this at the early stage of its business. Today’s startups are mostly IC design manufacturers. If they want to become "self-sufficient" companies like Intel and Samsung, they may need to spend billions of dollars.

After the wave of semiconductor industry integration in 2015-2016, the wave of semiconductor mergers and acquisitions in the past two years is gradually "cooling down", and large companies will be more cautious in their investment or acquisition of chip startups.

The core competitiveness of cloud AI chips lies in talents.

Judging from the more concerned cloud AI chip companies in the current market, their research teams are mostly industry veterans with more than ten years of experience in chip giants, and often have the experience of taking the lead in developing relevant successful products.

Both semiconductor giants and cross-border core-building technology giants are basically taking two paths. One is to invest in and acquire mature chip companies, and the other is to poach chip executives from other big companies.

Song Jiqiang, president of Intel Research Institute, once told Zhizhi that the future of AI chips must be diversified. Different kinds of products meet the requirements of different power consumption, size and price. AI is a marathon, and now this game has just begun.

At this stage, the vast majority of giants and entrepreneurs in the field of incoming cloud AI chips are playing innovative signs, including innovative architecture, storage technology and silicon light technology.

Due to the surge in demand for new computing resources to promote deep learning, many people think that this is a rare opportunity for start-ups to win funds from giants and investment institutions.

Although the number of players is increasing and the flags played tend to be diversified, at present, the innovative hardware that is really mass-produced is still limited. There are still many difficulties faced by cloud AI chips, such as Moore’s Law, which is common in computer architecture, and bottlenecks in semiconductor devices.

The process of developing chips may take several years, and most of the hardware is still under development or in the early test plan. Therefore, it is difficult to predict which enterprises will achieve the promised performance.

Generally speaking, the cloud AI chip market is gradually divided into three forces: semiconductor giants represented by NVIDIA and Intel, Chinese and American technology giants represented by Google and Huawei, and chip entrepreneurs represented by Cambrian and Groq. Among them, semiconductor giants and chip-creating enterprises focus on general-purpose chips, while cross-border core-making technology giants and AI-creating enterprises will not sell directly to the outside world for the time being.

From the application field, although the high energy consumption of GPU has been more and more criticized by the industry, due to its unparalleled parallel computing ability, there is no player who can compete with NVIDIA GPU in the field of cloud AI training. Players who challenge this field are mainly traditional chip giants and startups. Cross-border technology giants include Google, Baidu and Huawei, and the main architectures adopted are general GPU and ASIC.

In the field of cloud AI reasoning, which pays more attention to energy consumption, delay, cost and cost performance, there are relatively more players entering the game, and the advantages of FPGA and ASIC are relatively higher than GPU. Intel, which has a comprehensive AI chip layout, is gaining momentum, and other players are not far behind. Several major Internet giants in China and the United States have basically joined the battle, but the progress of chip research and development of some giants is still unknown.

With regard to improving the core-building strength, most semiconductor giants and technology giants have chosen the shortcut of investment, M&A and chip-digging, so as to be directly assisted by mature chip teams and quickly fill the vacancy of talents and business. For start-ups, there are basically two factors that are favored by the investment community-an experienced founding team and products with innovative technologies. From the perspective of the landing process, the pace of chip start-ups in China can rank among the top in the world.

At present, the vast majority of AI applications still rely on training and reasoning in the cloud. In the field of training, NVIDIA’s solid ecosystem is still an unshakable mountain, and in the field of reasoning, it is even more competitive. As AI is more widely spread to all walks of life, the cloud AI chip market will gain more room for growth, but this market may not accommodate so many players. Capital, device bottlenecks, architectural innovation, adapting to rapidly changing AI algorithms and building an ecosystem are all difficult problems facing these enterprises. What is the AI chip form that is completely suitable for cloud training and reasoning has not yet reached a unified conclusion.

When you ask a young man why he doesn’t like sports and what his pain points are, he will probably get the answer: he can’t find a partner.

Running, skipping and swimming can be done by one person, but there is always a lack of fun. Sports in which two or more people must travel in a row are always full of uncertainty. Either no one can be invited, or several people with different technical levels reluctantly accommodate each other.

There are always endless excuses, but when young people see that the cholesterol value in the physical examination report is increasing year by year, they have to take action and look around for sports games that can be played by themselves without being too boring.

(Source/"My Savage Girlfriend")

Squash is simply tailor-made for them, the most cost-effective choice.

Squash is similar to tennis, but the court usually only needs a semi-closed indoor rectangular space with a length of 9.75 meters and a width of 6.4 meters, and it takes only a few steps to run the whole court.

Other ball games need to practice all kinds of footwork, grip and hitting posture for a long time, but even zero-based squash students can perfectly explore all the abilities of entertaining themselves within 2 hours;

Squash courts give players a sense of privacy. (Source/vision china)

In the era when everyone claimed to be I, as long as he spent tens of dollars, he could confidently enjoy an entire air-conditioned space alone. Those scenes where the ball next door flies over and needs to be picked up and asked if it is possible to make up the field have never belonged to squash.

The unique advantages of venue and game mechanism have cleared away all the excuses of young people for not exercising. At the moment when the squash door is closed, the young people who have been sitting in the office for a long time finally have the opportunity to release themselves without scruple, draw the opposite wall full of black spots, and then leave smartly.

For ordinary people, it is often in TVB dramas and Korean dramas that they first come into contact with the concept of squash.

In the play, the executives go straight to the squash court after work, change their sportswear, swing against the wall and play ball. During the break, several people talk to each other, and maybe some cooperation has been reached.

TVB gold collar standard: playing squash. (Source/"Forensic Pioneer")

Now, while the executives are talking in Kan Kan in the squash court, almost all the venues next to them have become the best places for young white collars who have just entered the workplace to start sports.

After graduating from college and returning to work in her hometown, Sally can find few friends to go out with. Colleagues either have families or work overtime. Looking at each other’s tired eyes, Sally was even embarrassed to send out an invitation to play ball together.

When searching on social media how to solve her sedentary "pear shape", Sally accidentally discovered this new gadget of squash.

Hammamy v Sherbini 2023 Qatar Classic Final. (Source /@5#Squash)

The rules of squash are not very similar to what we understand as "duet": in a semi-closed room, two players hit the ball against the wall in front of them, and the opponent needs to catch the ball that bounces off the wall and hit it again.

When there is no opponent, the addition of the wall allows the player to control the rhythm of hitting the ball by himself, and it is most comfortable and relaxed without accommodating other people’s movements.

Social terrorists are most afraid of the sudden "guidance" of strangers during sports, but squash requires a closed space, and the court must be closed when using it. As long as the door is closed, no one can tell whether judge’s own actions are standard.

"It doesn’t matter even if he says it. Anyway, in the glass room, I can only hear my own panting."

Nowadays, when searching for "squash", the key word is often "social fear". (Source/@ 么么么么么么么)

Coach Mo, who has more than 20 years’ experience in squash teaching, often meets golfers who play alone when waiting at the squash court. Although they are familiar faces who have played in the same venue for many years, they will still avoid the crowd. "Some people don’t even want to have normal language communication with us at all."

However, coach Mo understands this very well. After all, when he started playing squash, he took a fancy to its "at any time, one person can play" characteristics.

"You can go to the squash court at 3 o’clock in the morning to enjoy 7 pieces of land, or you can take turns to play in a field with five or six friends. Squash, that’s it,’ I can e’. "

Free from social leisure, although a person’s squash can relax his mood, the black ball with irregular high-speed beating in such a large venue can still touch the players’ nerves at all times. After several rounds, people who "mess up" at ordinary times will sweat profusely and their hearts will accelerate.

Squash can fly at a speed of 150 kilometers per hour. (Source/WeChat official account @ Hangzhou 19th Asian Games)

After the first solo squash experience, Sally rested for a whole week, but she still made an appointment for a squash experience class two weeks later happily rubbing her sore thigh muscles, hoping to try the "complete form" of the sport.

Unfortunately, Sally lost interest after playing a few rounds with strangers. Although everyone is a beginner, when she can’t catch the ball sent by the other side and is forced to run all over the court to pick it up, she will still be numb with embarrassment.

"I still have the pleasure of hitting the wall when I play. When I play with others, I only feel that the wall is hitting me."

Unlike Sally, a beginner, Dee, who has been playing squash for 22 years, thinks that squash is more like a social and brain-burning game.

Dee fell in love with squash when he was accidentally brought into contact with his classmates in high school 20 years ago.

Beginners who play ball once a week spend half a week to relieve sore muscles, but for squash fanatics, it’s easy to make appointments twice a week, and even spare time to participate in competitions all over the country.

(Source/"My Savage Girlfriend")

Nowadays, Dee has less and less entertainment time with friends, but more and more social clubs, and the golfers have basically become "middle class" and "executives" in the relative sense, such as foreign consuls and brand owners.

In business, another popular social sport is golf, but the effect of this relatively static and non-sweating sport is not so direct.

It is impossible for decent executives to rush to play basketball together. Holding a golf club with a sense of distance, taking into account safety and antagonism, squash seems to be the best choice for sports and social interaction.

"Although they won’t take care of my business, playing ball is indeed the lowest cost way to get closer."

There are few conflicts on the squash court, and everyone is polite. (source/worm idea)

Sally thinks that every time she waves a racquet, it’s like slapping a face on the wall with a small ball, which is a great pleasure to vent her anger.

For old players, every stroke means constant calculation and thinking from footwork to ball speed and trajectory, which not only does not decompress, but also burns the brain abnormally.

But this does not mean that old players can’t get emotional value by "smashing the wall". In the process of playing squash, Dee’s most fulfilling moment is the moment when the ball is "smashed".

The ball used in squash is a hollow rubber ball filled with inert gas. Whenever the ball is hit, the inert gas in the ball expands due to friction and the speed of the ball becomes faster and faster.

Double yellow rubber balls used in squash. (source/worm idea)

When the speed reaches the limit of the surface rubber, the squash will burst open at the moment when it hits the wall, and the inert gas inside will "hiss" and the rubber will cross a short parabola in the air and slide to the ground like a breath.

Like collecting medals, Dee will collect the smashed balls: "It’s like a friend who has been with me for a long time suddenly left, and I want to cherish its company."

The pursuit of technology has always been the ideal of hardcore sports players, and "decompression" is the ultimate reason why many sports can be persisted by most people for a long time.

Different from other sports, most students who seek a coach to learn squash do not want to hone their skills, but hope to entertain themselves and relieve their pressure after learning the basic movements.

"After all, this kind of sport, which is difficult to break the ball, wall and racket, is most suitable for venting negative emotions." Coach Mo said.

One-to-many squash experience class is a popular entry method at present. (Source/Respondent for the picture)

On the court, coach Mo has met students from 4 to 80 years old, but the main consumer group of squash is still young people. Among them, foreign trade, media, accounting, construction, finance, public relations, lawyers and other occupations accounted for the majority.

Young people engaged in these industries have a relatively long work cycle, and there may be complications at any time in the process. Therefore, there is always a string in their minds, and they dare not relax completely.

On the squash court, the ball path is unpredictable, and all attention needs to be mobilized in starting, hitting and dodging. High-intensity exercise is enough to make nervous young people forget the pressure of work for a short time and concentrate on the audio-visual enjoyment of the 360-degree all-round beating sound in the field.

From the muffled sound of hitting the ball at the beginning, to the blasting sound produced by learning to use physical strength to the rhythmic sound of hitting the ball continuously … The subtle change of the sound of the ball hitting the wall every time is the most beautiful decompression music for them.

After playing ball and getting high, skills seem to be less important. (Source/Shepherd photo)

What’s more, the amount of exercise in squash can be regarded as rich and frugal. Punching the wall for 15 minutes is enough to cover the consumption of a sedentary office for a week, and slowly hitting the wall can just reach the critical point of fat consumption in aerobic exercise.

I’m already very tired at work. If you want to vent your pressure after work, you’d better choose a way to make yourself more comfortable.

Dee has never regarded squash as an "aristocratic sport". On the contrary, this kind of sport, which is sun-free, air-conditioned, unaffected by the weather and has a low entry threshold, has a wide mass base.

After all, compared with the expenses of old-fashioned "middle-class sports" in the public impression, such as golf, fencing and equestrian, squash that can be played for tens of dollars is simply cabbage price.

More and more teenagers are playing squash. (Source/Respondent for the picture)

Coach Mo, who has worked for many years, believes that the biggest reason why squash is still a niche in the world is the space limitation.

A standard squash court has requirements for lighting, floor and air conditioning, and each court can only accommodate two people at a time, which is not a common sports basic configuration.

Nowadays, when more and more players value the sociality and entertainment of squash, the old players are both happy and tangled. Where there is demand, there will be supply. It is foreseeable that there will be more and more squash venues in the future.

Perhaps, as Sally said, "Before squash courts really become popular, whether you can continue to play ball may not depend on whether you have time, but whether you have hand speed."

Every week, coach Mo organizes a "novice game" with his golfers. (Source/Respondent for the picture)

Ending：

Around 1830, the aristocratic students of Harrow College, a famous British aristocratic school, were tired of the boring living on campus and invented an indoor sport of hitting the ball against the wall, which was the origin of squash.

Up to now, squash is still inseparable from the labels such as "elite sports", "middle class" and "minority", which seem to be far from the lives of ordinary people.

In fact, "tolerance" is the key word of squash. Beginners can entertain themselves, and hardcore players also have fierce play. On the squash court, everyone can find their own comfort zone.

Dee describes the feeling of being in a squash court with the feeling of being wrapped. (Source/vision china)

Most importantly, after closing the door, the player can immediately immerse himself into another world from the senses. Compared with the real life where hard work may not necessarily lead to results, the world of squash is so disciplined.

There, as long as you pay close attention to the direction of the ball and try to hit it on the wall, the sense of coolness will certainly accumulate.

Only then can the young people who are tired of life really find the pure happiness that belongs to sports without any label.

Acknowledgement: Coach Mo of Guangzhou Squash Club

Writing and herding sheep

Original title: "It’s hard for a young man who smells like a class not to like playing squash"

Read the original text