IDC: Baidu Smart Cloud tops AI public cloud ranking for the fourth time

IDC's latest "China AI Cloud Service Market 2020 Annual Research" report shows that in China's AI public cloud service market, Baidu Smart Cloud ranks first in China in terms of market share. In the segmented fields, Baidu Smart Cloud has ranked first in the NLP public cloud and image and video public cloud market shares for three consecutive times, and has achieved the first place in the face and body recognition public cloud market share for two consecutive times.

It is worth noting that this is the fourth time that Baidu Smart Cloud has topped the AI ​​public cloud list since IDC released the fourth phase of China's AI cloud service market.

In this regard, IDC evaluated that Baidu Smart Cloud still ranks first in the AI ​​public cloud service market share. AI applications are blooming across the board and being implemented in multiple industries.

As enterprises continue to transform digitally, AI capabilities are becoming the main factor in choosing cloud service providers. Therefore, cloud service providers on the market are gradually emphasizing the word "intelligence". However, in such a fiercely competitive AI Cloud market, Baidu has won the championship four times. This makes people curious about how strong Baidu Intelligent Cloud is and what is its unique understanding of enterprise AI applications?

For enterprise digitalization, "going to the cloud" has long become a must, because going to the cloud can not only reduce computing costs, but more importantly, it can aggregate scattered data and open up business processes, thereby better using AI technology to reshape the business.

As the first step in digital transformation, moving to the cloud is only the foundation of digital transformation. What is more important is to use artificial intelligence technology to enable enterprises to complete digital and intelligent upgrades.

However, as enterprises' digital transformation continues to deepen, their requirements for AI technology are becoming more pragmatic. The first priority is to be able to embed it into existing production factors and to continuously create value.

According to a joint survey conducted by Baidu and Boston Consulting Group in 2018, 86% of the market demand is customized AI demand. From 2017 to 2020, customized AI models increased more than 6 times, which means that more and more companies are beginning to develop models according to their own needs.

Due to the surge in customized AI models, companies have also put forward higher requirements for AI development capabilities. On the one hand, there is a demand for AI model accuracy and AI development efficiency. On the other hand, the demand for AI computing resources has also shown explosive growth.

In the future, as AI is deeply applied in more scenarios, AI will have to process more and more data, the models will become larger and larger, and the customization will become deeper and deeper. Therefore, the demand for computing power will become higher and higher: the amount of calculation will become larger and larger, the accuracy requirements will become higher and higher, and the processing time requirements will become higher and higher. As a result, the computing cost will become higher and higher.

At the same time, AI development involves a series of tasks such as data collection, data preprocessing, data labeling, model training, model evaluation and model deployment. The process is long, complex and involves many steps. Each step may become an efficiency bottleneck and may increase the cost of AI development.

Therefore, for enterprises, maximizing efficiency under the premise of controllable costs is an issue that must be focused on in AI development. These costs include computing costs and development costs, and efficiency includes data quality, model accuracy, processing effects, and result accuracy.

In this regard, Baidu Smart Cloud recently released the "Cloud Intelligence Integration" Technology and Application Analysis Series White Paper - Enterprise Development, which believes that the AI ​​development platform that adapts to the future must have the following three capabilities:

1. In terms of computing resources, it is necessary to connect and manage AI computing resources with high performance, high utilization rate, and high cost performance;

2. In terms of R&D efficiency, we need to provide full-process optimization and development efficiency tuning;

3. Provide overall solutions based on the actual business scenarios and customized needs of enterprises.

Among the many cloud service providers in China, Baidu's understanding of AI is unmatched.

In early 2010, Baidu took the lead in establishing a "Natural Language Processing Department" to serve the relevant intelligent needs generated by search engines; then it established a department called the "Multimedia Department" in voice and image technology, and laid out key capabilities such as machine vision and intelligent image processing that are now popular; almost at the same time, Baidu also finalized its strategy of long-term investment in knowledge graph technology, and soon built the world's largest Chinese knowledge graph network.

Since then, Baidu has never given up its investment in AI technology research and development. Over the past 10 years, Baidu has invested nearly 100 billion yuan in AI technology research and development. It is precisely the continuous investment that has made Baidu clearly realize that for the digital transformation of enterprises, "cloud" and "intelligence" are not just a simple superposition relationship.

In 2016, Baidu Smart Cloud proposed the "ABC Trinity" strategy, integrating AI, cloud computing and big data.

Since that year, Baidu has been promoting the integration of cloud and AI. At the Baidu Cloud Intelligence Summit on May 18, 2020, Baidu CTO Wang Haifeng landed on the development route of "based on cloud computing, using AI as a starting point, and focusing on important tracks", further establishing the development strategy of "cloud and intelligence integration".

Thanks to Baidu's accumulated experience over the years, Baidu Smart Cloud's "cloud-intelligence integration" is not simply adding AI components to cloud computing, nor is it cloud-based AI technology services. Instead, it is a deep integration of cloud and AI, using the cloud to carry AI and using AI to define the cloud.

In terms of "cloud", Baidu Smart Cloud has integrated AI elements into the "AI development infrastructure (equivalent to traditional IaaS)". The specially tuned AI development infrastructure can function with higher performance and high utilization; and supports various heterogeneous computing powers, including CPU, GPU and customized chips in various fields; at the same time, the computing deployment location has also been distributed, with a large amount of computing distributed in large data centers, various edge terminals, as well as smart parks and smart factories, making Baidu Smart Cloud the most suitable cloud for running AI.

In terms of "intelligence", although there are various AI technology service providers and corresponding cloud services, they often do not have native cloud capabilities. Baidu Smart Cloud's "AI Development Platform" is integrated into the AI ​​native cloud computing architecture. At the same time, PaddlePaddle has become China's first open source, fully functional deep learning platform, which has attracted 3.2 million developers, an increase of nearly 70% compared to one year ago; it serves 120,000 enterprises and institutions.

AI-defined cloud and cloud-native AI make Baidu Smart Cloud the only "cloud-intelligence integrated" cloud computing platform and AI development platform, forming full-stack AI development capabilities to better meet the needs of enterprise AI development.

Not long ago, Baidu released its Q1 financial report for fiscal year 2021, showing that Baidu Smart Cloud's revenue reached 2.8 billion yuan in the first quarter, a significant increase of 55% year-on-year, and the growth rate is at the leading level among all cloud computing platforms.

Although its scale cannot yet be compared with other cloud vendors with revenue of tens of billions, it can be seen from the continued growth rate that Baidu Smart Cloud's potential has been expanded due to its joint layout with AI.

On this basis, Baidu Smart Cloud continues to build connections with industry partners, and while building AI infrastructure in the digital economy era, it also understands the industry better.

In addition, Baidu has also successively launched industrial Internet infrastructure construction projects in Guiyang, Guizhou, Jinan, Shandong, Chongqing and other places. While laying the foundation for the new AI infrastructure in the local area, it also works with local governments and enterprises to "build nests to attract talents."

For Baidu, this is not only beneficial for laying a solid foundation for independent control, but also for continuously consolidating the establishment of a certain cooperation foundation with the partner ecosystem, which is also the key measure for the phased development of Baidu Smart Cloud.