ECARX and Arm sign MoU to develop next-generation AI computing platforms
ECARX and Arm have signed a non-binding Memorandum of Understanding to jointly explore and develop next-generation high-performance AI computing platforms for mobility and wider cross-industry uses. The two companies announced the agreement on Nov. 2, 2026, in London.
- ECARX and Arm signed a non-binding MoU on Nov. 2, 2026 to explore next-generation high-performance AI computing platforms.
- Joint work covers architecture definition, performance optimisation, software, functional safety, toolchain and cybersecurity validation, and ecosystem build-out.
- Platforms are aimed at mobility and cross-industry physical AI, with configurations for China and global markets.
- The deal builds on collaboration via SiEngine, whose 7nm SE1000 SoCs have been in mass production since March 2023.
- A 5nm SE2000 SoC platform is under development, according to the release.
What the MoU covers
According to the announcement, the MoU sets a framework for exploring advanced solution development, including joint research and development and product definition. ECARX, which describes itself as a leading global automotive intelligence business, and Arm plan to carry out high-level joint co-engineering work.
The activities named in the release span several areas:
- system-level architecture definition
- performance optimisation
- software enablement
- functional safety
- toolchain and cybersecurity validation
- ecosystem build-out
The partners will also look at tailored configurations optimised for deployment in China and global markets. These would address use cases in mobility as well as broader cross-industry physical AI applications. The companies say the combined effort will unlock meaningful advances in performance, power efficiency and scalable deployment for global customers.
A relationship that builds on SiEngine
The agreement follows an existing collaboration between the two companies. The release points to SiEngine, which it calls a leading automotive semiconductor company. According to the announcement, SiEngine has delivered its 7nm SE1000 platform of automotive-grade SoCs to global OEMs, with millions of units produced and deployed since mass production began in March 2023. Development of a next-generation 5nm SE2000 SoC platform is also underway.
Under the new MoU, Arm will support ECARX in bringing to market highly optimised compute solutions built for demanding physical AI workloads, the release says.
What the executives said
Ziyu Shen, Founder and Chief Executive Officer of ECARX, said the collaboration comes against an industry shift toward AI-defined vehicles. In his words, it "accelerates our work to bring forward globally ready high performance physical-AI computing platforms." He added that the teams will now advance joint technical roadmap definition and co-engineering.
Drew Henry, Executive Vice President, Physical AI at Arm, said physical AI is "one of the most complex systems engineering challenges our industry has taken on." He said it requires compute, software and safety to work together from the ground up, and that combining Arm's compute platform with ECARX's automotive expertise offers an opportunity to explore new approaches to efficient high-performance computing across AI-defined vehicles and other autonomous systems.
About ECARX and what the MoU means
ECARX is listed on Nasdaq under the ticker ECX and is headquartered in London. According to the company, its technology is deployed across more than 12 million vehicles worldwide, and it works with 18 global automakers and 28 vehicle brands. It was founded in 2017, listed on Nasdaq in 2022, and has more than 1,400 employees across 13 major international locations.
The MoU is non-binding, so it sets out an intention to explore joint work rather than a committed product plan. The release gives no timelines, pricing or performance figures for the planned platforms. For now, the announcement signals that ECARX and Arm intend to define a joint technical roadmap for high-performance computing aimed at AI-defined vehicles and other physical AI uses.
