Energy Efficient Open-Source RISC-V Processor for Automotive Workloads

Abstract

Modern vehicles integrate a growing number of electronic control units to perform key tasks, such as driver assistance, smart braking, and steer by wire. This presents a challenge for the transition towards battery electric vehicles as any power consumed by embedded systems cannot be used for driving, directly reducing vehicle range. As a result, energy efficiency is of central concern as automotive software grows more complex. Another factor is ease of development, with closed systems requiring expensive and specialized software for development, making them expensive and difficult to maintain. Moreover, processors traditionally used in the automotive space are proprietary which enables vendor lock in. Exploring alternative processor architectures has great potential for meeting the evolving requirements. The open and flexible nature of RISC-V makes it particularly well-suited to address the performance, safety, and energy-efficiency challenges of modern automotive systems. The instruction set being open allows anyone to design a compatible processor which can enable competition and result in cheaper and more efficient designs. Additionally, the design can be tailored to the specific performance and power requirements of the application. Finally, software development can leverage open-source tools, reducing the need for expensive licenses. This thesis investigates the suitability of the NOEL-V, an open-source RISC-V processor, for use in automotive systems. It is compared to a commercial automotive ECU based on Infineon TriCore TC399XP in terms of performance and power consumption by porting a climate system to RISC-V and evaluating it on an FPGA implementation of NOEL-V. The performance is assessed through cycle count and schedulability of real-time tasks, while power consumption on a 45 nm process is estimated using EDA tools. For the ECU, the performance metrics of the TriCore are collected on the real hardware through Lauterbach debug tools, and power consumption is estimated indirectly via the system-level power increase during workload execution. The results show that although the NOEL-V platform does not achieve the same raw performance as the TC399XP, it is sufficient for the climate application and requires fewer clock cycles per instruction. The project also identifies major power hotspots in the RISC-V processor under the target workload, and proposes directions on optimizing power dissipation for future RISC-V-based automotive processors.