Set micro-mesh news, the new year to open new hope, the new space to carry a new dream .This is the beginning of the first set of micro-network readers to write a message when the New Year .In the Chinese New Year opening year, the semiconductor industry For example, Yangtze River Storage is negotiating with Apple to purchase the former Nand flash chip, and another example is Wang Haili, a former vice president of Beijing Weijiage who persevered in the second pioneering work 18 months later.
At the end of 2005, Wang Haili, who is going to graduated from Tsinghua University with a Ph.D. in computer science, joined a newly established Sino-foreign joint venture, Yahoo! Technology, to start the research and development of China-made FPGA chips. In 2010, with support from the Beijing Municipal Government, Also converted identity and changed its name to 'Jingwei Ya grid'.
In the beginning, because Xilinx and Altera accounted for more than 90% of the market, and intellectual property barriers built with nearly 9,000 patents with companies such as Lattice and Microsem, they were blocked to a great extent later Who pursued the road.Jingwei Ya grid initially because of patent restrictions almost impossible, had another way - bought the United States CSwitch, acquired the patent, coupled with some expired patents of some FPGA vendors, and eventually successfully developed Tile framework of their own.
However, FPGA is a product with large fund consumption, long development time and long payback period. With the interruption of continuous capital investment, the related funds are not put in place in time, and the company lacks the rhythm of financing and the problem of 'playing the best role of food and drink' In front of the team in May 2016, set a micro-grid disclosure of Beijing micro-Accor encountered difficulties, two months of arrears, people have to go to the floor empty situation.After Jingwei Ya Ge responded: 'The company took place in business Some difficulties, but there is no case of company collapse, the core management team still hold the post, the company is looking for reorganization.
At that time, Wang Hai Li, the vice president, was one of the eight remaining employees, but the reality was not as good as they had imagined. During the period, the bankruptcy of the company led to an abnormal financing problem and staff labor dispute, property and Supplier recovery and other difficult issues, the state of business deteriorating.
In seeking the consent of the shareholders of the original company premise, we finally decided to start a second business, staying in the eight people in June 2017 set up 'micro-Beijing (Qili Technology)', Wang Hai invested 2000000 yuan, chairman and concurrently General Manager The original company continues to maintain the status quo, at the same time, all its products, technologies and other exclusive license to the company now use the team hopes to find ways to solve the problems faced by the original company.
Born "Beijing micro-Qi (Qi Li Technology) ', Continuation of FPGA Chuixin
Through the acquisition of CSwitch, based on a number of FPGA vendors partially expired patents based on independent research and development, after the Beijing micro-Accord applied for a total of more than 200 patents, and has more than 90 patents have been granted.
After reorganization, the team of "Beijing Micropower (Qili Technology)", after obtaining the authorization of over a hundred patents and proprietary technologies (including international patents) from Jingwei Yage and secondary development rights, provided the traditional FPGA chip At the same time, we further aimed at the application needs of artificial intelligence and emerging applications, opening up a new generation of Ai programmable logic controller (AI in FPGA), Heterogeneous Programmable Accelerator (HPA), embedded programmable eFPGA IP research and development plan.
Funding is the first thorny issue facing the new company.It is fortunate that less than 2 months after the establishment of the new company, it has got nearly 7 million yuan of angel round of financing, with shareholders of Tsinghua-based investment funds and strategic investors, For the follow-up development of the company has laid a solid foundation.This round of financing led by Thai fund leader, Mizuki Fund, Zhen cloud venture, Tyco source with the vote.
The financial difficulties have been solved, the next is the talent problem.The original company's core technical team is the first member of the forward design of FPGA chips, Wang Haili understand that these people have to be invited again, and now the company employees have been from the beginning Eight people to the development of more than 30 people, including four returned overseas experts, a Beijing Distinguished Specialist and the sea poly people, a national "thousands of people plan" expert consultants, the majority of the team has more than 20 years of work experience in the industry.
As for what products to do, Wang Haili said in a salon blocked by Thai funds that Beijing Micro Microsystems is currently using some simple AI applications with its own small-capacity FPGA products.
In AI applications, FPGAs have the advantages of high speed, low latency, low power consumption, etc., and have strong I / O expansibility (I / O will be the key once the AI enters the product phase) and will have in the future In addition, due to the programmable, configurable and re-programmable FPGAs, FPGAs can take advantage of FPGAs to continuously update their products to quickly capture the market.
He said that based on the above market analysis and FPGA's own unique technological advantages, in addition to continuing to sell the original FPGA chip, the company is developing a new generation of AiPGA chips for applications such as artificial intelligence / intelligent manufacturing. Heterogeneous Computing HPA chip, embedded programmable eFPGA IP core, the product market will cover the cloud server, Internet of Things, consumer smart terminals and national communications / industrial / power / military / medical and other core infrastructure.
In rebuilding the server room, restoring data and recruit older employees, the team has already set up two new product lines: 'Hercules' chip and 'Avatar' chip, the former will be used in consumer emerging video-related fields, the latter The main attempt 16/12 nanometer new technology to further enhance the existing FPGA's power and energy consumption ratio.
According to pencil reports, 'Qi Li Technology' is now producing 40-nanometer FPGA chip, the next two years will be 28 nm and 16/12 nm advanced process nodes to achieve a breakthrough.'s Existing customers more than 40 companies in 2017 total Achieve sales of more than 10 million yuan.The team in January of this year again completed the tens of millions of Pre-A round of financing, the lead investor for the Hankook fund.
According to Wang Haili revealed that new products have also entered the research and development stage, the company expects sales this year will reach 20 million yuan. 'Experience of the past two years, I realized the spark of the stars can start a prairie fire.'
At the same time, domestic FPGA manufacturers are also gradually showing a hundred schools of thought contend .It has been engaged in the reverse design, quite marketable in military products, the country Microelectronics; there is a technical team from Lattice's core team of Xi'an Polycrystalline; and be backed by violet, and with the micro-electronics is quite similar to Tongchuang core.
The difficult road of domestic FPGA, whether it is the beginning of the new micro-Beijing, or other domestic manufacturers, domestic independent FPGA enterprise's future and development are noteworthy.
2. ZTE released "White Paper on AI Artificial Intelligence Network"
Xinhua News Agency, Beijing, February 17 (Reporter Gao Kang) ZTE recently released a whitepaper titled "AI Artificial Intelligence Network Intelligence" to the industry. The white paper focuses on the artificial intelligence of communication networks. Combined with ZTE's latest research and practice in network intelligence , Comprehensively expounded the future intelligent network architecture, solutions and scenarios with the vision of 'network autonomy, foresight of the future, mobility on demand and smart operation'.
You Yan, vice president of ZTE, said ZTE believes that intelligent is the future trend of network development, network operation and operation and maintenance model will undergo fundamental changes in the future, from the current people-driven model of people-based governance, and gradually self-driven network-based The evolution of the autonomous model.An intelligent network will be through the network data, business data, user data and other multi-dimensional data perception, based on big data, big calculation and big algorithm three basic capabilities to achieve a high degree of autonomy.
ZTE is closely watching the evolution of telecom operators' digital transformation and network intelligence. Based on the uSmart I nsight platform, ZTE Corporation offers a full range of big data, big power and big algorithm capabilities to help telecom operators build a future-oriented Intelligent Network. "You Yan said that in the field of artificial intelligence, ZTE Corporation has carried out in-depth cooperation with dozens of operators and partners in the world to jointly explore and promote the application, sharing and popularization of artificial intelligence + network.
3.Workday set up 250 million US dollars fund, focusing on artificial intelligence, blockchain and other cutting-edge technology;
Workday Ventures (Nasdaq: WDAY) today announced the launch of Workday Ventures, a $ 250 million fund. Workday Ventures is the company's strategic investment unit that focuses on investing in early stage companies and working with them to develop applications Emerging technologies in the enterprise such as AI and machine learning, blockchain and augmented reality, virtual reality.
Workday executives Leighanne Levensaler and Mark Peek have been named managing directors and co-heads of the fund, while Leighanne will continue to be Senior Vice President of Workday's Enterprise Strategy Mark is an industry veteran with more than 30 years Financial and operating experience, previously co-president and CFO of Workday, having held leadership positions in companies including Amazon and VMware.
It is reported that this 250 million US dollars fund is set up after Workday announced that it will open its cloud platform to create new opportunities for its client and partner ecosystem to expand the core capabilities of Workday.In addition to funding, Companies that Workday Ventures invest in will also gain Workday's rich experience, market knowledge and global reach in enterprise cloud applications, including:
Gain guidance from Workday product executives and executives who guide how companies invest in building and delivering technology on a large scale, and creating sustainable business models.
Connect with Workday customers who are some of the largest and most disruptive brands in the world with more than 26 million users.
Gain practical guidance at the technical level, including with Workday data scientists, architects, and engineers.
Aneel Bhusri, co-founder and CEO of Workday, said: "The launch of Workday Ventures strengthens our commitment to the forefront of enterprise innovation." 4. Google new research: the development of human joyful artificial intelligence;
SAN FRANCISCO, February 17 morning news, Google's team is using daily human behavior to guide the machine decision-making, the process does not require any programming.
Researchers have developed a web-based application that shows people artificial-inspired paintings such as kittens and rhinos and records their reactions through the camera, which is a positive signal if people see them laughing.If people seem to Very confused, then it is a negative signal.
After gathering all of this feedback, the Google team re-trained the AI system and focused on the 'good' example, with the end result that artificial intelligence can create the paint better.
Although the experiment looks simple, the Google team points out that this ensures that AI is safer in the long term if AI is capable of adapting to human social behavior and can learn from human facial expressions and body language It's better to do this kind of behavior to make human happier.
Of course, the reverse is also true: Artificial Intelligence learns what behavior fears or disgusts humans, and MIT researchers are exploring the field, but Google is focusing on the positive side.
Google is not the only company to do that, and companies like Affectiva are exploring technologies that have artificial intelligence in mind and technologies that allow machines to better understand human facial expressions and social behaviors to make the machine more responsive to human needs, such as Affectiva Study how to make driverless cars understand that the driver is switching to autopilot mode.
Google's team pointed out: 'A human-driven artificial intelligence is not likely to be against the good of humanity.'