Jul. 30, 2025 –
The modern vehicle is no longer just a mechanical machine, but a dynamic computing platform defined by software and evolving toward an era of AI-defined intelligence that learns, adapts, and improves over time. For automakers and suppliers, this shift is reshaping everything from product roadmaps to procurement strategy.
Since the 2021 global chip shortage, which exposed vulnerabilities across the supply chain, one question has taken center stage: Should companies bring more control in-house, or partner more deeply across the value chain? At first glance, vertical integration promises tighter control, and some OEMs are already exploring custom chips and proprietary software stacks. However, this can come at a high cost – in capital, time, and talent.
The pace of innovation and complexity of future AI-defined vehicles suggest another path: strategic collaboration, or what some now call “coopetition.”
Modern vehicles need to do far more than ever before. A centralized compute platform simultaneously handles in-vehicle infotainment (IVI), advanced driver assistance systems (ADAS), power management, and more. These workloads vary in criticality and need to run in real-time, often without cloud connectivity. In this context, both performance and efficiency are critical.
Not only do vehicles need to operate with tight power and thermal constraints, but they also need to reach stringent safety requirements. As a result, technologies like chiplets are gaining ground because they enable more modular, scalable systems without the need to reinvent every layer.
However, the success of these advancements hinges on shared standards. Without open interfaces and cooperative design frameworks, the risk of fragmentation increases.
The appeal of custom silicon is clear. If done right, it can optimize compute performance, reduce reliance on third parties, and offer a degree of differentiation.