Cloud computing has become the biggest trend recognized in the world. Gartner research report shows that as of the end of 2015, the global cloud computing market has reached 175 billion US dollars, and it is expected to exceed 300 billion US dollars in 2019. Domestic BAT giants are also eyeing this market.
This time Baidu announced a brand upgrade, upgrading "Baidu Cloud Computing" to "Baidu Cloud". Unlike Alibaba and Tencent, Baidu is best at artificial intelligence, and this time Baidu has also invested its weight in cloud computing in the field of artificial intelligence.
Baidu President Zhang Yaqin said that in the future, with the exponential growth of big data and the breakthrough progress of artificial intelligence, "cloud computing, big data and artificial intelligence" will become the key to future development. The next scene of the Internet will be artificial intelligence. Begin to dominate, and all walks of life will also rely on "intelligence" to complete their own transformation, and Baidu Cloud has also carried out a comprehensive upgrade for this purpose.
In fact, driven by artificial intelligence, future cloud computing is entering a new intelligent field, which mainly shows three major trends.
First, big data surges, and computing power becomes the new focus of cloud computing
There is no such thing as cloud computing in the traditional era, because the problem that needs to be solved in the early days is storage. Thousands of corporate websites need to be connected to the Internet, and these companies only need hosting services.
But the mobile era is a critical turning point. With the advent of the mobile era, the data left by users in enterprise products has begun to surge. Compared with the PC era, it can be described as astronomical, and E-commerce, logistics, medical care, education, marketing, finance and many other industries have been fully affected. These companies increasingly need to use user data to adjust their products and improve planning, and have a huge demand for computing power. .
Previous cloud computing solved the storage problem, but it did not solve the problem of how enterprises handle big data. Therefore, how to efficiently process big data will be a new focus in the future. When storage capacity is no longer important, computing power is becoming a new pursuit of future enterprises.
The "Tiansuan" released by Baidu Cloud this time is in response to this demand. Baidu itself started by relying on computing power. Search requires a large amount of data to be calculated from the beginning, and its average daily response to search More than 6 billion times, covering 97.5 Internet users across the country, LBS has more than 30 billion daily requests, and daily speech recognition requests have exceeded 100 million times. These data have given Baidu's computing power powerful training, constantly pushing its own capabilities.
The opening of "Tiansuan" is to share the 16 years of accumulation with the whole society, so that every industry can solve the problem of big data utilization most efficiently.
Second, user interaction methods have begun to diversify, making it difficult for enterprises to cope with them
In the early PC era, due to the inconvenience of operating equipment, the interaction between users and machines was only through text, but Smartphones in the mobile era have brought great portability, and users' request methods have begun to take many forms, including picture requests, voice requests, video requests, and many other forms.
Only by seizing these new forms of interaction can enterprises seize the future. Therefore, artificial intelligence technology is needed to respond to changes in user scenarios and help their users use their products more conveniently and efficiently, regardless of whether the user Both voice requests and picture requests can be responded to correctly, but in fact the entry barrier to this field is extremely high, and it is impossible for most companies to set up separate departments for research and development, so third parties providing artificial intelligence services have become key.
Facing the above new user interaction scenarios, Baidu Cloud released "Sky Image". The Sky Image function can provide developers with cloud services for multimedia processing, storage, and distribution of videos, pictures, documents, etc., and also Additional services such as "anti-pornography services", "visual effects", "face recognition", "text recognition" and more are provided. In addition, Baidu Ecosystem also provides huge traffic support to "Tianxiang", allowing companies to gain huge traffic exposure.
Corresponding to changing user scenarios, Baidu has long been the focus of its layout. Taking speech recognition and image recognition as examples, Baidu’s accuracy in speech recognition is as high as 97%, and its DeepSpeech2 deep speech learning is also It was named one of the top ten breakthrough technologies by the MIT Review, alongside aerospace technology and biotechnology, and was the only Chinese company selected. In terms of image recognition, Baidu's DeepImage can identify image content. The accuracy of LWF face recognition reaches 99.86, and the accuracy of text OCR is also 90.
At present, Baidu has also fully opened up its voice and image technologies. In addition to accessing "Sky Image", developers can also directly use Baidu's open platform for voice recognition and image recognition to easily complete their own tasks. High-threshold technical implementation that cannot be completed.
Third, the rise of the Internet of Things and the comprehensive evolution of cloud computing to artificial intelligence
Following German Industry 4.0, my country also proposed the Made in China 2025 plan in 2014, and intelligent engineering was also officially On the topic, in 2025, key manufacturing industries will be fully intelligent and achieve unified intelligent management.
Cloud computing of the Internet of Things is different from other cloud computing. Its focus is not on storage and hosting. It requires a standardized management rule so that devices can have unified access, unified scheduling, unified detection, etc. etc., and all of this relies on artificial intelligence technology, which means that in terms of cloud computing for the Internet of Things, traditional cloud computing is no longer competent. It not only needs to be combined with artificial intelligence, but also needs to evolve itself into artificial intelligence.
The "Tiangong" released by Baidu Cloud this time is a comprehensive layout for the future Internet of Things era, and Baidu itself has also made considerable moves in the Internet of Things. Unmanned vehicles can be said to be The master of the entire Internet of Things involves many aspects of top-level software and hardware technologies, and Baidu has listed unmanned vehicles as its focus.
Baidu has used the most advanced fully autonomous driving of unmanned vehicles as its standard from the beginning, using the world's most advanced sensors and other equipment, and combining it with its own cutting-edge voice technology, image recognition technology, map technology, and deep learning. technology to realize the operation of unmanned vehicles, and has also received great support from the government. It has obtained pilot permits from Wuhu, China, and California, the United States. It is not only the first company in Asia to make unmanned vehicles, but also the only one. Companies recognized by both governments.
Unmanned vehicles are a comprehensive integration of top cutting-edge artificial intelligence technologies, and "Tiangong" opens these cutting-edge artificial intelligence technologies so that more IoT developers can obtain low-cost, high-quality For efficient solutions, Tiangong currently has cooperation cases with IoT companies such as Dengyun, Hudao Information, Puao Cloud, Zhixiang Technology, and IoT World, and will also attract more and more top-notch companies in the future. developers.
Conclusion:
The future is the era of data, the era of multiple requests from users, and the era of Internet of Things devices. Traditional managed cloud computing will not be able to do the job, and cloud computing will also be comprehensive. Evolving towards artificial intelligence, there will be no cloud computing but artificial intelligence in the future, and this trend has already begun.