On April 23 at the Shanghai Auto Show, Intel unveiled its second-generation AI-enhanced Software-Defined Vehicle (SDV) SoC. This SoC, designed on a chiplet architecture, enhances generative and multimodal AI performance up to 10 times over the previous generation. Additionally, the graphics performance sees an improvement of up to 3 times, facilitating a richer Human-Machine Interface (HMI) experience. The SoC supports 12 camera channels, enhancing camera input and image processing capabilities.
Intel also announced collaborations with several industry leaders.
In partnership with Black Sesame Technologies, Intel launched a cockpit-driving fusion platform that integrates Intel’s AI-enhanced SDV SoC with Black Sesame’s Huashan A2000 and Wudang C1200 chip families, catering to automotive needs from L2+ to L4 driving levels. This platform offers an open, flexible, and highly expandable design, allowing manufacturers to fit one design into multiple models and simplifying the development process.
The reference design for the cockpit-driving fusion platform is set for release in the second quarter of 2025 with plans for mass production.
HyperAI and Intel have established a strategic partnership to introduce an in-vehicle pure edge GUI agent that provides users with offline voice command understanding, contextual memory, personalized service recommendations, and screen operations. In complex conversational scenarios, the agent smoothly interprets language structure and context, understanding natural language commands, making human-machine interaction more seamless and natural.
Moreover, Intel announced a collaboration with South Korean semiconductor company BOS Semiconductors. Leveraging BOS’s automotive AI accelerator chiplet SoC Eagle-N, Intel’s AI-enhanced software-defined SoC will offer automotive manufacturers higher performance and more powerful AI solutions.
This article is a translation by AI of a Chinese report from 42HOW. If you have any questions about it, please email bd@42how.com.