Domestic 5nm Longying 2 Chip Launches with 200TOPS AI Performance

At the 2026 Beijing International Auto Show, a significant breakthrough was achieved in high-performance automotive-grade chips. Xincheng Technology officially unveiled its independently developed 5-nanometer automotive-grade AI cockpit chip, the "Longyang 2." This launch represents a crucial advancement for domestic high-end vehicle chips in both advanced process technology and cross-domain integration capabilities.
The "Longyang 2" chip, the centerpiece of this announcement, features exceptionally powerful specifications. Built on the industry-leading 5-nanometer process, it delivers AI computing power of up to 200 TOPS and natively supports multi-modal large language models with over 7 billion parameters. This computational strength not only readily handles current mainstream intelligent cockpit demands but also provides ample performance headroom for future, more complex human-machine interaction and in-vehicle AI applications.
Architecturally, the "Longyang 2" employs a forward-looking "flexible architecture." This highly adaptable design caters to the evolution of central computing platforms, from entry-level to flagship vehicle models. A key innovation is its full-scenario "cockpit and driving integration" capability. The chip integrates dedicated vehicle control processing units and security islands. Utilizing physical isolation technology, it can safely process both cockpit infotainment and advanced driver-assistance tasks on a single hardware platform, significantly reducing vehicle electrical/electronic architecture complexity.
Founded in 2018 through a joint investment from Ecarx Technology and Arm China, Xincheng Technology leverages deep industry expertise to focus on high-performance automotive-grade chips. The "Longyang 2" is scheduled to begin vehicle integration in the first quarter of 2027. Its subsequent mass production and delivery are expected to provide automakers with a more competitive domestic computing solution, accelerating the automotive industry's shift toward the central computing era.
Related article
Pentium 4 Revival: 20-Year-Old CPU Runs Meta Llama 3 Large Model
Recently, the YouTube tech channel Fully Buffered carried out an impressive and hardcore experiment: successfully running Meta's latest Llama 3.2 3B large model on the Pentium 4 641 processor, a chip released in 2006.This test forced modern artificia
Hangzhou Shangcheng District Launches Zhejiang's First AIGC Audio-Visual 'Golden Ten Measures', 5 Billion Industry Fund
On the 16th, the AIGC Audio-Visual Industry Innovation Ecosystem Conference took place in Hangzhou's Shangcheng District. During the event, the province unveiled its first dedicated policy for the AIGC audio-visual industry—"The Golden Ten." This pol
MIIT Seeks Public Feedback on 121 Industry Standards, Including AI Model Context Protocol
China's Ministry of Industry and Information Technology has officially released a notice seeking public feedback on 121 industry standardization projects, including the "Application Security Requirements for the Artificial Intelligence Security Gover
Related Special Topic Recommendations
Comments (0)
0/500

At the 2026 Beijing International Auto Show, a significant breakthrough was achieved in high-performance automotive-grade chips. Xincheng Technology officially unveiled its independently developed 5-nanometer automotive-grade AI cockpit chip, the "Longyang 2." This launch represents a crucial advancement for domestic high-end vehicle chips in both advanced process technology and cross-domain integration capabilities.
The "Longyang 2" chip, the centerpiece of this announcement, features exceptionally powerful specifications. Built on the industry-leading 5-nanometer process, it delivers AI computing power of up to 200 TOPS and natively supports multi-modal large language models with over 7 billion parameters. This computational strength not only readily handles current mainstream intelligent cockpit demands but also provides ample performance headroom for future, more complex human-machine interaction and in-vehicle AI applications.
Architecturally, the "Longyang 2" employs a forward-looking "flexible architecture." This highly adaptable design caters to the evolution of central computing platforms, from entry-level to flagship vehicle models. A key innovation is its full-scenario "cockpit and driving integration" capability. The chip integrates dedicated vehicle control processing units and security islands. Utilizing physical isolation technology, it can safely process both cockpit infotainment and advanced driver-assistance tasks on a single hardware platform, significantly reducing vehicle electrical/electronic architecture complexity.
Founded in 2018 through a joint investment from Ecarx Technology and Arm China, Xincheng Technology leverages deep industry expertise to focus on high-performance automotive-grade chips. The "Longyang 2" is scheduled to begin vehicle integration in the first quarter of 2027. Its subsequent mass production and delivery are expected to provide automakers with a more competitive domestic computing solution, accelerating the automotive industry's shift toward the central computing era.
Pentium 4 Revival: 20-Year-Old CPU Runs Meta Llama 3 Large Model
Recently, the YouTube tech channel Fully Buffered carried out an impressive and hardcore experiment: successfully running Meta's latest Llama 3.2 3B large model on the Pentium 4 641 processor, a chip released in 2006.This test forced modern artificia
Hangzhou Shangcheng District Launches Zhejiang's First AIGC Audio-Visual 'Golden Ten Measures', 5 Billion Industry Fund
On the 16th, the AIGC Audio-Visual Industry Innovation Ecosystem Conference took place in Hangzhou's Shangcheng District. During the event, the province unveiled its first dedicated policy for the AIGC audio-visual industry—"The Golden Ten." This pol
MIIT Seeks Public Feedback on 121 Industry Standards, Including AI Model Context Protocol
China's Ministry of Industry and Information Technology has officially released a notice seeking public feedback on 121 industry standardization projects, including the "Application Security Requirements for the Artificial Intelligence Security Gover





Home






