RISC-V Summit 2022

The automotive industry is at an inflection point where major disruptions are happening. One of them is how the current compute architecture, where each function is attributed to an ECU, moves to a centralised architecture akin to a human brain. The central compute architecture uses data from different sensors around the vehicle, and AI inference to help take the right decisions. The requirements of these compute systems are expanding. The main ones being it should be able to work with any type and amount of data, be scalable in compute performance and should be efficient and simple to deploy. As the requirements for the system drive greater computation and efficiency, it is expected that the architecture becomes increasingly heterogenous which will require flexible programming models and software frameworks. This talk will give a good understanding where the automotive industry’s challenges in terms of compute are and how the next generation solutions will hinge on different forms of heterogeneous compute solutions.