Recently, the Beijing Auto Show has been held in full swing, with the strong rise of new energy vehicles and intelligent networked vehicles, the intelligent cockpit has gained a lot of attention and has become a must for everyone. The experience of intelligent cockpits has been evolving for a long time, and as AI fully empowers thousands of industries, generative AI is defining a new cockpit experience. In the context of technologies such as end-side large model on-board and computing power evolution, the market opportunities and competition pattern of the intelligent cockpit industry are being reshaped, and the world's leading semiconductor companies represented by MediaTek are accelerating this change.
The 3nm flagship cockpit platform was unveiled, demonstrating the strength of the strongest automotive computing chips
During the auto show, the debut of the 3nm Dimensity automotive cockpit platform CT-X1 caused a lot of waves in the industry, and was considered to be the number one natural enemy of the friend 8295.
The CT-X1 adopts the most advanced 3nm process in the industry, which refreshes the new standard of the flagship cockpit chip and provides confidence for the top computing power in the manufacturing process. At the same time, this 3nm Dimensity flagship cockpit chip also has an advanced heat dissipation design, which helps to fully release the computing power of the chip, and it is officially announced that its computing power will exceed that of similar flagships in the market in 2023 by more than 20%, becoming the computing power ceiling of a new generation of cockpit chips.
According to the news broke by Antutu, MediaTek demonstrated the intelligent cockpit solution based on CT-Y1 at the Dimensity automotive platform communication meeting held, as a 4nm automotive-grade chip, the measured Antutu car machine version has a running score of more than 1.07 million, tied with Snapdragon 8295, and reached the top flagship level. The CT-X1 created by the 3nm process was unveiled at the same time, with 30% stronger performance than the Snapdragon 8295 (this is a very conservative number from the perspective of architecture), and the AI computing power is 4-5 times stronger, fully demonstrating the strength of the strongest automotive computing chip.
In terms of AI artificial intelligence, CT-X1 leads the large model on the car, supports 13 billion parameter AI large language models, much higher than industry standards, makes the cockpit smarter, and can run a variety of mainstream large language models (LLMs) and AI drawing functions (Stable Diffusion) in the car, providing advanced AI safety and entertainment applications such as in-vehicle voice assistant based on 3D graphical interface, rich multi-screen interaction and display technology, and driving alertness monitoring. End-to-device AI not only improves safety, but also has the advantages of high response speed and low latency, which brings unprecedented experience innovation to intelligent cockpits.
In addition to the CT-X1, the Dimensity automotive cockpit platforms CT-Y1 and CT-Y0 were also unveiled, using a 4nm process, with strong computing power, and also supporting advanced end-to-end generative AI technology, supporting 7 billion parameter AI large language models, and also having the ability to generate AI images within 1 second. With the help of the Dimensity automotive cockpit platform, automakers can upgrade the intelligent experience from flagship (CT-X1) to high-end (CT-Y1 and CT-Y0) models, meeting the needs of different positioning models for high computing power and strong AI.
Compared with the mainstream flagship cockpit chips in the market, the latest CT-X1 shows industry-leading strength in terms of manufacturing process, computing power, and AI capabilities, and can be called the strongest automotive computing chip.
Leading the automotive industry into a new era of "AI-defined cockpit".
The development of the current intelligent cockpit to the future intelligent cockpit will open a new era of "AI-defined cockpit".
The AI of traditional intelligent cockpits is generally composed of the device side and the cloud, and the device side generally only supports small-scale models with less than 1 billion parameters, which obviously cannot meet the needs of the rapidly developing intelligent cockpit experience. In the future intelligent cockpit, it will support large models with more than 7 billion parameters to achieve strong intelligence on the device and side. In addition, the future intelligent cockpit also has a full-time online large-model automotive assistant to provide users with low-latency and safer interactive services, and also supports end-to-side multi-modal technology to provide richer automotive AI functions.
With its powerful AI capabilities, the Dimensity automotive cockpit platform will bring exciting application experiences in driving experience, social networking, office, entertainment and more. For example, the full-time online large-model car assistant can interact with the user's voice, gestures, actions and emotions, and in terms of entertainment, it can automatically edit and share the scenery inside and outside the car during self-driving trips.
Through the combination of devices and clouds, it can help the intelligent cockpit provide personalized restaurant and navigation route recommendations, as well as applications such as health monitoring and tracking, intuitive voice interaction, and car manuals. Not only that, but it can also provide personalized news recommendations and summaries, of course, music recommendations based on user preferences, and even interactive games based on voice, touch screen or vision, and immersive in-cabin innovation experiences such as picture and story generation.
Whether it is for industry trends or consumer decisions, intelligent cockpits are playing an increasingly important role, and MediaTek integrates generative AI technology into the Dimensity automotive cockpit platform, leading the way in the end-to-end implementation of generative AI through strong performance, innovative applications, and the integration of mobile phones and automotive ecosystems. It is not difficult to predict that the track of intelligent cockpit will undergo revolutionary iterations, with high computing power on the device side as the base, leading generative AI technology and rich application ecology as the drive, MediaTek has become more and more an important participant and promoter of the era of AI-defined cockpit.
Join forces with the Innovation Bureau, MediaTek + Nvidia > Qualcomm
Some industry insiders said that the rapid success of MediaTek's cockpit platform is inseparable from another giant in the industry, NVIDIA. Combined with the AI leader NVIDIA, the Dimensity automotive cockpit platform holds the two trump cards of the strongest AI and the strongest computing power for automobiles, which makes MediaTek expected to break the existing pattern and make MediaTek + NVIDIA > Qualcomm a new highlight of the industry, and the in-depth combination of AI and computing technology is an inevitable trend in the development of the smart car industry.
As an old player in the automotive industry, MediaTek Dimensity automotive platform has maintained a continuous growth momentum in the market, among which the global market shipment of Dimensity automotive cockpit platform has exceeded 20 million sets, the Dimensity automotive connectivity platform has been adopted by the world's leading automobile manufacturers, the shipment and market share of key components such as Dimensity automotive satellite navigation system and power management chip continue to grow, and the progress of Dimensity automotive driving platform is also very smooth. Nvidia is a well-known pioneer and leader in artificial intelligence and computing, and the two companies announced a collaboration in May last year that leverages the strengths of their respective automotive portfolios and combines the best of both companies' capabilities to develop automotive SoCs with NVIDIA GPU chiplets and NVIDIA AI and graphics computing IP to deliver superior solutions for the next generation of smart cars.
For the first time in a year, a series of AI-enabled Dimensity automotive cockpit platform SoCs were unveiled at NVIDIA GTC, featuring the C-X1, C-Y1, C-M1 and C-V1 cockpit platform chipsets that integrate the advanced Armv9-A architecture with AI computing and NVIDIA powered by NVIDIA's next-generation GPUs RTX graphics processing technology and deep learning capabilities create a highly competitive AI cockpit experience, bringing users of all levels of vehicles a far superior cabin experience.
In the mobile phone market, you can always trust the performance of MediaTek CPUs, combined with NVIDIA's top AI and GPU graphics technology, the two sides have jointly created a powerful and feature-rich AI cockpit solution, which is enough to be proud of the entire cockpit chip industry. At the same time, as a leading player in the field of AI, the two parties will work together in automotive AI to lead the sustainable development of the AI-defined cockpit era. Under the trend of cockpit integration in the future, the ecosystem cooperation between the two parties based on high-computing power AI and high-capability AI is undoubtedly more efficient and is the optimal solution for the upgrading of the intelligent cockpit industry.
Nowadays, new energy vehicles are leading the global wave, China's new energy vehicles are rising strongly, and in the long run, the entire automobile market contains huge upward potential. Facing the general trend of AI-defined automobiles, the in-depth cooperation of technology giants has shown strong market influence, which poses a great challenge to MediaTek's opponents, and it is the fierce product and technology competition that is the driving force behind the continuous innovation and development of the automotive industry.