Research on Automotive Cloud Services: As Dedicated Automotive Cloud Platforms Are Launched, the Market Enters A Phase of Differentiated Competition
1. The exponentially increasing amount of vehicle data makes cloud migration an inevitable choice.
From the perspective of companies, the goals of digital transformation are to digitize all elements of the whole process throughout the full life cycle of vehicles, including R&D, production, sale, operation, and after-sales service; upload the data in the local servers and computer rooms of automakers to the cloud; connect the data channels of each link to gradually realize the integrated management of data in the whole industry chain, and the cloud-pipe-terminal integrated real-time interconnection; and build service operation models that span the full life cycle of users to enhance the connections between upstream and downstream partners in the industry and create greater value.
In terms of products, vehicle intelligence and connectivity are booming. For example, starting from L2, every time the autonomous driving functions evolve to a higher level, the consumption of cloud infrastructure platforms, applications, and services will rise by an order of magnitude. As high-level autonomous driving comes into mass production, the number of vehicle sensors and the amount of data multiply, making it difficult for local processing to meet the requirements. Cloud migration thus will be the best choice.
Automakers spend tens of millions of yuan every year building cloud services, which gives a big boost to the market. In 2022, China's automotive cloud service market was valued at over RMB15 billion, and it is expected to sustain the growth rate of 30-40% in the next five years.
2. As dedicated automotive cloud platforms are launched, differentiated competition becomes crucial.
In 2021, ByteDance announced the "ByteDance Auto Cloud", which will provide cloud services in four segments: Digital Marketing, Intelligent Cockpit, Autonomous Driving, and Vehicle Services. In 2022, Tencent Intelligent Cloud Cloud, Baidu Auto Cloud, and Alibaba Auto Cloud became available. All the five giants (BATHD), i.e., Baidu, Alibaba, Tencent, Huawei and Douyin have stepped in the market, and the competition in automotive cloud services built on exclusive automotive cloud has become fiercer.
The service scope of each automotive cloud is much of a muchness, generally covering R&D, manufacture, marketing, and supply chain. The support for R&D is concentrated in the fields of autonomous driving, intelligent cockpit, telematics, and “three electrics” (battery, motor and ECU). How to gain differentiated competitive edges in the competition therefore has become the key to success for companies.
3. The differentiated competitive edges in cloud services are mainly built from two aspects: basic resource layer services and upper-layer R&D tool chains.
In terms of basic resource layer, supercomputing centers are an important indicator for assessing service capabilities, and Alibaba and Baidu are the first to deploy.
In August 2022, Alibaba Cloud launched the two intelligent supercomputing centers located in Zhangbei County and Ulanqab, with total compute of 15 EFLOPS (15 exascale floating-point operations per second). At the same time, Alibaba Cloud also introduced the "Apsara Intelligent Computing Platform", an intelligent computing solution which opens up intelligent computing capabilities by way of "platform + intelligent computing center".
Following the five intelligent computing centers in Yangquan, Jinan, Fuzhou, Yancheng, and Tianjin, Baidu Cloud started construction of the Baidu AI Cloud-Shenyang Intelligent Computing Center in May 2023, a project with a land area of about 2.4 hectares, floor areas of 42,000 square meters, and the total planned computing power of 500P, 200P for Phase I. In the future, Baidu Shenyang Intelligent Computing Center will not only involve physical data center construction capabilities and intelligent computing infrastructure capabilities, but also comprehensive solutions for AI software and hardware ecosystem capabilities such as foundation models, supporting the computing tasks of companies in different business scenarios and meeting the industrial application requirements of foundation models in the era of intelligent computing.
With regard to R&D tool chains, cloud service providers are committed to creating "fully furnished" service experiences for users by offering "full-process" and "fully closed-loop" services.
In Tencent's autonomous driving cloud platform, virtual simulation has become a key link.
Huawei's autonomous driving cloud platform "Octopus" has built in a dataset with 20 million frame annotations, a library with 200,000 simulation scenes, a complete tool chain, and annotation algorithms, covering the full life cycle businesses such as autonomous driving data, models, training, simulation, and annotation, and helping automakers to build autonomous driving development capabilities on a "zero" basis.
Baidu makes a full-stack layout and enables a data closed loop by virtue of from chip (Kunlunxin), deep learning (PaddlePaddle) and training foundation model (ERNIE) to search (Baidu Search), cloud platform (Baidu AI Cloud), autonomous driving (Apollo) and intelligent connection (Xiaodu).
4. Under the multi-cloud strategy, the need of OEMs has changed from the pursuit of resources to efficiency.
With the in-depth migration to the cloud, the resource needs of OEMs for cloud migration have been overall met, and thus the underlying logic of the cloud strategy of companies has changed from the pursuit of resources to efficiency to finally improve their overall digitization capabilities in production and operation. In this process, OEMs are no longer tightly bound with some cloud platform, but implement a multi-cloud strategy where different business types are put on different cloud platforms.
Examples include:
Based on the "1+6+N" Geely Hybrid Cloud Platform co-built with Baidu, Geely works with Alibaba to build the Xingrui Intelligent Computing Center, and teams up with Tencent on telematics and security solutions.
FAW Group uses Huawei Cloud Stack to build hybrid cloud architecture, and also cooperates with Alibaba Cloud on intelligent manufacturing, digital marketing and other businesses.
Without a doubt, the multi-cloud strategy offers benefits. It can integrate the advantages of various cloud platforms, enable refined business deployment, and reduce costs for companies, and also helps automakers to gain the core initiative in building cloud platforms and avoid being puzzled by the “soul” dispute. Yet the challenges of the multi-cloud strategy are also unavoidable. How to allocate storage/computing power among multiple clouds, cross-cloud data synchronization’s dependency on bandwidth, and whether costs and network delays will have an impact are all urgent problems to be solved. Hence how to formulate a multi-cloud strategy is a problem for OEMs.
China Automotive Multimodal Interaction Development Research Report, 2023
China Automotive Multimodal Interaction Development Research Report, 2023 released by ResearchInChina combs through the interaction modes of mainstream cockpits, the application of interaction modes i...
Automotive Smart Surface Research Report, 2023
Market status: vehicle models with smart surfaces boom in 2023
From 2018 to 2023, there were an increasing number of models equipped with smart surfaces, up to 52,000 units in 2022 and 256,000 units ...
Passenger Car Intelligent Steering Industry Report, 2023
Passenger Car Intelligent Steering Industry Report, 2023 released by ResearchInChina combs through and studies the status quo of passenger car intelligent steering and the product layout of OEMs, supp...
Automotive High-precision Positioning Research Report, 2023-2024
Autonomous driving is rapidly advancing from highway NOA to urban NOA, and poses ever higher technical requirements for high-precision positioning, highlighting the following:
1. Higher accuracy: urb...
New Energy Vehicle Thermal Management System Research Report, 2023
Thermal management system research: the mass production of CO? heat pumps, integrated controllers and other innovative products accelerates
Thermal management of new energy vehicles coordinates the c...
Commercial Vehicle Intelligent Chassis Industry Report, 2023
Commercial Vehicle Intelligent Chassis Industry Report, 2023, released by ResearchInChina, combs through and researches status quo and related product layout of OEMs and suppliers, and predicts future...
Chinese Independent OEMs’ ADAS and Autonomous Driving Report, 2023
1. Wide adoption of NOA begins, and local brands grab market share.
According to ResearchInChina, from January to August 2023, joint venture brands accounted for 3.0% of installations of L2.5 and hi...
Passenger Car Radar Industry, 2022-2023
Passenger Car Radar Industry Research in 2023:?In 2023, over 20 million radars were installed, a year-on-year jump of 35%;?Driven by multiple factors such as driving-parking integration, NOA and L3, 5...
Automotive Audio System Industry Report, 2023
Technology development: personalized sound field technology iteration accelerates
From automotive radio to “host + amplifier + speaker + AVAS” mode, automotive audio system has passed through several...
China Intelligent Door Market Research Report, 2023
China Intelligent Door Market Research Report, 2023 released by ResearchInChina analyzes and studies the features, market status, OEMs’ layout, suppliers’ layout, and development trends of intelligent...
Automotive Infrared Night Vision System Research Report, 2023
According to the data from ResearchInChina, during 2022-2023, the installations of NVS (night vision system) in new passenger cars in China went up at first and then down. From January to July 2022, t...
New Energy Vehicle Electric Drive and Power Domain Industry Report, 2023
Electric drive and power domain research: electric drive assembly evolves to integration and domain control
To follow the development trend for electrified and lightweight vehicles, new energy vehic...
Automotive Software Business Models and Suppliers’ Layout Research Report, 2023
From the layout of automotive software products and solutions, it can be seen that intelligent vehicle software business models include IP, solutions and technical services, which are mainly charged i...
Automotive LiDAR Industry Report, 2023
In August 2021, Waymo discontinued its commercial LiDAR business.
In October 2022, Ibeo declared bankruptcy; in November, two listed companies, Velodyne and Ouster, confirmed their merger; and in Dec...
Automotive Power Supply (OBC+DC/DC+PDU) and Integrated Circuits (IC) Industry Report, 2023
Automotive power supply and IC: Chinese chips are promising in the evolution from physical integration to system integration
As the core component of a new energy vehicle, automotive power supply is ...
OEMs’ Model Planning Research Report, 2023-2025
OEMs’ Model Planning Research Report, 2023-2025, released by ResearchInChina, combs through model planning and features of Chinese independent brands, emerging carmakers, and joint venture brands in t...
Leading Foreign OEMs’ADAS and Autonomous Driving Report, 2023
Global automakers evolve to software-defined vehicles by upgrading EEAs.
Centralized electronic/electrical architectures (EEA) act as the hardware foundation to realize software-defined vehicles. At ...
Automotive AI Algorithm and Foundation Model Application Research Report, 2023
Large AI model research: NOA and foundation model facilitate a disruption in the ADAS industry.
Recently some events upset OEMs and small- and medium-sized ADAS companies, as the autonomous driving i...