At the Right Place, At the Right Time: How Baidu is Cementing its Position as a Leading AI Company

Baidu’s cloud computing center in Yangquan, Shanxi province. (Source: Baidu Research)

For two decades, Baidu Inc has been referred to as China’s Google. Now, with so much positive momentum surrounding the company, investors, analysts, businesses and consumers alike are gradually realizing that it is much more than a search engine.

The company made its debut on the Hong Kong Stock Exchange on Tuesday, with shares rising more than 1% in early trade from their issue price of HK$252 ($32.45).

While jumping into the race to develop cloud computing services and AI-powered technologies in the digital economy era, Baidu also experienced heightened demand during the Covid-19 pandemic, sparking increased growth in its non-advertising business – revenue grew at a rate of 52% year-on-year in the fourth quarter of 2020. The primary factor driving the recent growth is the company’s AI cloud business, which grew by 67% and reached an annualized run rate of $2 billion.

Analysts believe that these developments, together with the firm’s IPO in Hong Kong, serve as a clear indicator that Baidu is well-positioned as a leading AI company with a strong Internet foundation, capable of seizing the enormous market opportunities in cloud services, autonomous driving, smart transportation and other AI-driven businesses.

In the next few years, Baidu expects the growth rate of its non-advertising segment, which currently makes up 18% of its core revenues, to be about three times faster than online marketing. Given its commitment to AI, it is very likely to meet – or even exceed – that expectation.

Baidu’s Differentiated Cloud Solution

In China, the overall market size of public cloud services reached 1.09 billion yuan ($168 million) in the first half of 2020, a year-on-year increase of more than 100%, according to figures released by market research firm IDC. Among them, the IaaS (Infrastructure as a Service) market increased by 57.1% and PaaS (Platform as a Service), market increased by 64.4% year on year.

This significant growth is propelled by a range of factors: first, an increase in demand for digital transformation and online consumption in response to the ongoing global health crisis; second, accelerated construction and investment of cloud data centers across various industries; and third, continuous investment from the government.

With more demand for cloud services, investment in other technologies such as data integration, mobile platforms and applications, and most importantly – AI and IoT (Internet of Things) are expected to follow. At the same time, IDC researchers believe the main driving force underpinning the public cloud service market in the future will come from demand for rapid business response and agile delivery, shifting away from cloud resource supply and cost.

No doubt the Beijing-based company knew this when it introduced Baidu Cloud – a one-stop shop with advantages of differentiated AI solutions for both enterprise customers and the general public.

Baidu Co-founder and CEO Robin Li at the bell-ringing ceremony for the company’s secondary listing on the Hong Kong Stock Exchange on Tuesday morning. (Source: Baidu)

In 2016, Baidu became the first Chinese tech giant to propose the “ABC” technology stack – AI + big data + cloud computing – to seek better integration among the three fields. Other companies soon followed suit, with Alibaba introducing its “cloud computing + IoT + artificial intelligence” strategy in 2018 and Huawei Cloud launching its “Inclusive AI” solution also in the same year.

Baidu CEO Robin Li believes that the development of cloud computing and the accumulation of big data are the foundation for the rapid advancement of AI, as well as the key to achieving practical breakthroughs. In turn, improvements in big data and AI will also expand cloud computing applications to empower and digitally transform enterprises.

Leveraging its unique, “full-stack” competitive edge in AI, algorithms, open platforms and developer ecosystems, Baidu has delivered a strong execution on its cloud and intelligence integration strategy.

For the third time, Baidu took top spot in China’s AI public cloud market, according to the latest IDC report, which stated that the company has made momentous achievements in areas of facial and body recognition, image and video processing, and natural language processing (NLP).

For many of Baidu’s enterprise clients, delving into cloud services is no longer just a fad, but a necessity. They realize that unless they innovate through AI, they will be left behind entirely.

“As AI is regarded by more and more people as the new engine driving economic growth, enterprise-level customers are recognizing a real demand for AI-backed cloud computing services to solve practical problems,” tech analyst Alter wrote in Forbes China.

“There is an urgent need to establish an architecture system that properly integrates AI and cloud computing in order to transform existing delivery methods of AI capabilities,” he said.

This is where Baidu enters. As of October 2020, the company had filed a nation-leading total of 9,364 AI-related patent applications in China. Meanwhile, PaddlePaddle, the largest open source deep learning framework in the country, has an accumulated 17,692 pull requests. Baidu’s data centers cover more than ten regions including Beijing, Baoding, Suzhou, Nanjing, Guangzhou, Yangquan, Xi’an, Wuhan and Hong Kong. Its AI open platform built on Baidu Cloud has been offering more than 270 core AI capabilities.

More importantly, during its ABC Summit in August last year, Baidu announced a new generation high-performance physical server named Baidu Taihang, along with six new products including the Gaia Database, four upgraded application platforms and an AI-Native cloud computing architecture that combines all of the above hardware and software into a one-stop service.

Baidu announcing its upgraded AI-Native cloud computing architecture in August last year. (Source: fonow)

“Only with forceful computing power can large-scale data be processed and more advanced algorithms can be applied,” Baidu CTO Wang Haifeng said during the event.

“Traditional cloud computing is all about computer system resources. But through adjustments to the system architecture, Baidu Cloud has systematically integrated cloud computing and artificial intelligence, furthering its advancement towards ‘All in AI’,” Alter said.

Successful Case Studies: Call Centers, Finance Services and Smart Transportation

“Competition in the field of cloud computing has never been a purely technical battle,” wrote Alter. “Marketing strategies, planning, responsiveness and coordination are also indispensable aspects. The seamless integration between cloud and AI in Baidu’s organizational structure allows it to react quickly to the market and come up with even better solutions.”

What originally was a differentiation advantage for cloud vendors to provide AI capabilities has now become a bottom-line prerequisite to attract new customers, he added.

Baidu’s AI call center, powered by Baidu Brain, once had telecom operators as its early customers but now the service has also been adopted by airlines, financial services, energy and automobile sectors to become an industry standard.

In 2019, the company also assisted Shanghai Pudong Development Bank in developing an assistant virtual bank teller for its mobile app, as well as an online terminal that can sense mood changes in customers and help with decision-making. The two parties signed a deal again in last July, vowing to deepen cooperation in AI and fintech.

Joining hands with the Economic and Technological Development Zone in Guiyang, a major national data center hub in Guizhou province, Baidu provided its AI PaaS platform to more than 400 local enterprises across ten industries.

Its AI PaaS platform was also deployed last year in a project with China National Building Materials (CNBM), the parent of 13 listed companies, to provide intelligent logistics and industrial autonomous driving technologies. With Baidu, it will be easier for CNBM to improve the routing efficiency of its logistics vehicles and track their whereabouts.

Working with China Pacific Insurance Company, Baidu developed an AI auto damage assessment platform that can detect fraud and process insurance claims with a 90% accuracy rate.

It also embarked on the IoT journey by offering smart home technologies such as DuerOS, a self-developed AI-backed voice assistant solution and DuerOS-powered smart devices.

Baidu’s Kunlun II chip. (Source: Baidu Research)

Its next-generation AI processor, Kunlun II, is scheduled begin mass production later this year. At present, more than 20,000 Kunlun I chips have been deployed to support Baidu’s search engine and Baidu Cloud partners since they launched in 2018, empowering industrial manufacturing, smart cities, smart transportation and other fields.

While Apollo, Baidu’s autonomous driving technology, often takes the spotlight in demonstrating the firm’s prowess in intelligent transportation, it’s important to note that its cloud computing infrastructure serves as a vital digital foundation in the formula.

“Given Baidu’s significant leadership in autonomous testing miles at over 80% market share, it is well positioned as the dominant operating system for both the smart transportation network and in-car technology,” Mizuho Securities Asia Vice President Wei Fang said in a report.

SEE ALSO: Baidu Plans to Raise up to $3.6 Billion in Hong Kong Secondary Listing

Baidu has so far won government contracts in major cities such as Guangzhou, Beijing, Shanghai and Chongqing. Its latest project, announced on Wednesday, involves the construction of a 105-million-yuan 5G intelligent driving project based in Chengdu’s Singapore-Sichuan Hi-Tech Innovation Park.

In the specific case of Guangzhou, Baidu has completed the first phase of a smart transportation project involving the deployment of Apollo V2X (vehicle-to-everything) smart road infrastructure and Apollo Mobility as a Service (MaaS) during the Chinese New Year holiday period.

Covering 102 intersections in Guangzhou’s Huangpu district and utilizing a fleet of five different model types including Apollo Robotaxis and Robobuses, the initiative allowed local commuters to hail smart transport services on demand using either Baidu Maps or the Apollo Go app.

Baidu has completed the first phase of a smart transportation project in Guangzhou. (Source: Baidu)

At the same time, Baidu launched its full-stack ACE Transportation Engine (autonomous driving, connected road, efficient mobility), which is essentially a transportation cloud solution. Through processing traffic information from Apollo MaaS, V2X, DuerOS connected vehicles and Baidu Maps, it can assist agencies with traffic management and is expected to improve traffic efficiency by 15% to 30%.

“The reason why Baidu possesses such a crushing advantage in the AI cloud field is not only due to its deep technical capabilities in AI, but also its keen insight in the cloud computing field,” analyst Alter said.

The Nasdaq-listed company is gearing up for new engines of long-term growth, analyst Zhang Shule told China Daily, as it intends to reduce reliance on its search advertising business.

Baidu has made ambitious promises for its cloud business. The company claimed that by 2030, it would have deployed five million intelligent cloud servers that are seven times more powerful than the combined computing strength of the world’s top 500 supercomputers.

As companies seek to gain a competitive advantage from cloud computing, such as higher efficiency, greater flexibility and a boost in performance, Baidu’s cloud and AI integrated solution is fundamentally changing companies’ traditional way of thinking about IT resources, while at the same time unlocking significant potential in its non-advertising business for years to come.