[Shanghai, China, April 29, 2021] Huawei launched the MDC 810, a new-generation computing platform for intelligent driving, at an event prior to Auto Shanghai 2021. The MDC 810 — now ready for mass production — has the industry's highest computing power, supporting 400 trillion operations per second (TOPS). It meets the functional safety requirements of Automotive Safety Integrity Level D (ASIL D). Underpinned by the MDC Core software (which includes the autonomous driving operating system [AOS] and the intelligent vehicle control operating system [VOS]) and a complete toolchain, the MDC 810 is aimed at high-level autonomous driving applications, such as traffic jam pilot (TJP), highway pilot (HWP), and automated valet parking (AVP).
▲Li Zhenya, General Manager of the MDC Product Department at Huawei, unveils the MDC 810.
"Currently, the intelligent driving industry is gaining momentum, racing towards mass production," said Li Zhenya, General Manager of the MDC Product Department under Huawei's Intelligent Automotive Solution BU. "Huawei is dedicated to building an open and standardized MDC. Based on its solid experience and achievements in hardware, software, and security engineering, Huawei has built the most powerful computing platform for intelligent driving — MDC — accelerating the mass production of intelligent vehicles."
Li explained that before a computing platform for intelligent computing could be mass produced, a host of challenges had to be resolved, spanning engineering of hardware, software, and security. This is a complicated, systematic process, and requires the conversion of technologies into engineering capabilities.
In hardware engineering, vehicles must be made resistant to dust, water, condensation, and electromagnetic interference. In software engineering, any centralized computing platform must enable both separation and sharing of bottom-layer hardware resources, and its own deterministic latency must be minimized. In security engineering, it is vital to deliver deterministic functional safety compliant to ASIL D, and take into account the Safety of the Intended Functionality (SOTIF). A comprehensive network security architecture and defense capabilities, which require years of effort to develop, are also crucial. Generally, it takes two to three years before a computing platform for intelligent driving can be ready for mass production.
Huawei has been building its capacity for mass production since it announced its first MDC product in 2018. The company has designed the MDC in strict compliance with automotive-grade requirements. During development and verification stages, over 200 design verification tests were conducted in labs, in addition to summer and winter drive tests. To maximize the performance of the MDC, Huawei has optimized everything, from chipsets, platforms, and operating systems to operators and the AUTOSAR adaptive platform/classic platform, bringing the MDC to the forefront of the industry in terms of latency. A comprehensive security architecture and relevant capabilities are in place to secure MDC hardware, platforms, access, and applications.
Huawei's R&D roadmap for its MDC highlights a unified hardware architecture, a unified software platform, and the company's intention to offer a whole series of MDC. The MDC series, now including the MDC 210, MDC 300F, MDC 610, and MDC 810, are all powered by a leading platform and security architecture, and provide diverse sensor interfaces. Their computing power ranges from 48 TOPS to over 400 TOPS. The evolvable MDC Core software (including AOS and VOS) supports forward compatibility and long-term evolution, helping customers to protect their investment and accumulate capabilities. At present, the MDC products can support various use cases involving passenger vehicles, commercial vehicles, and special-purpose vehicles with levels of autonomy ranging from L2+ to L5.
HUAWEI MDC is an open, standardized platform aiming to boost the growth of the intelligent driving industry. Huawei positions itself as an engine for industry development and catalyst for integration across the value chain, with the goal of tilling the soil for this industry. Huawei has been working closely with vendors of sensors, actuators, and application algorithms to build such technology ecosystems based on the MDC. Through the open computing platform, Huawei unites the efforts of eco-partners to accelerate construction of the intelligent driving ecosystem, stepping into the mass production phase of intelligent vehicles.
About Huawei IAS BU
Huawei is a leading global provider of information and communications technology (ICT) infrastructure and smart devices. We are committed to bringing digital to every person, home and organization for a fully connected, intelligent world.
As electric vehicles become more intelligent, the automotive industry is relying more on ICT technologies such as AI, digital power, and cloud services. This industry will surely become the new engine for development in the intelligent world. In May 2019, Huawei established an Intelligent Automotive Solution (IAS) BU through which we are bringing more than 30 years of ICT expertise to the automotive industry. We are working with partners to drive the transition from internal combustion vehicles to intelligent connected vehicles. Huawei is positioning itself as a supplier of new components for intelligent connected vehicles. By focusing on computing, operating systems, networks, cloud, and other ICT technologies, we have created five major solutions that cover intelligent driving, intelligent connectivity, intelligent cockpits, mPower, and intelligent vehicle cloud services. Together with our industry partners, we're working to bring digital to every vehicle and build a better world for intelligent travel.