Beyond Automotive: Chiplets as the Foundation for Autonomous Edge Platforms

Edge High Performance Compute is becoming the “brain” of autonomous systems, enabling the rapid growth of L2+ and future autonomous mobility capabilities. However, the automotive industry faces a fundamental dilemma: compute requirements are increasing exponentially, driven by AI and advanced perception workloads, while the global automotive market remains relatively flat and cannot absorb the rising costs associated with leading-edge semiconductor technologies. 

Imec believes interoperable chiplet architectures provide the path forward. By decomposing complex system-on-chip designs into modular building blocks, chiplets enable heterogeneous integration of compute, AI, safety, and memory functions, each optimized on the most appropriate technology node. This approach improves scalability, yield, cost efficiency, supply-chain resilience, and platform reuse while supporting the transition toward software-defined vehicles. 

To accelerate adoption, imec is bringing together complementary initiatives across the chiplet value chain. The Automotive Chiplet Forum (ACF) aligns industry stakeholders around open, interoperable chiplet standards, while the Automotive Chiplet Program (ACP) advances pre-competitive research and technology development. Complementing these efforts, the new Chiplet Acceleration Center (CAC) at imec Germany supports design enablement, prototyping, and the path toward industrial deployment. Together, they help de-risk development, advance quality and reliability, and accelerate the transition from chiplet innovation to real-world automotive applications. 

Building on the success of ACP, imec is expanding its vision beyond automotive with the launch of the Autonomous Edge Chiplet Program (AECP), addressing shared compute needs across robotics, humanoids, aerospace, and security applications. Complemented by the development of a reference autonomous edge platform, this cross-industry strategy aims to accelerate innovation, reduce costs, strengthen resilience, and establish interoperable chiplets as the foundation for next-generation autonomous systems.