Renesas Electronics Corporation has developed processor technologies for automotive SoCs employed in applications such as ADAS and AD systems that strive to optimise performance and power efficiency while maintaining a high level of functional safety. The company’s successful development includes a CNN hardware accelerator core that produces a world-class combination of deep learning performance of 60.4 TOPS and a power efficiency of 13.8 TOPS/W; sophisticated safety mechanisms for fast detection of and response to random hardware failures. This makes it feasible to produce a highly power-efficient detection mechanism with a high failure detection rate; and a mechanism that enables software tasks with various safety levels to work in parallel on the SoC without interfering with each other, whereby reinforcing functional safety for ASIL D control. These technologies have been used in the company’s latest R-Car V3U automotive SoC.
The company presented these accomplishments at the International Solid-State Circuits Conference 2021 (ISSCC 2021), which took place 13 to 22 February 2021.
Applications including next-generation ADAS and AD systems need exceptional deep learning performance of 60 TOPS or even 120 TOPS alongside power efficiency. Also, since signal processing from object identification to the issuing of control instructions composes the bulk of the processing load in AD systems, delivering the functional safety equivalent to ASIL D – the most stringent safety level defined in the ISO 26262 automotive safety standard – is a pressing issue. The company has developed new technologies to satisfy these demands, including a hardware accelerator that gives outstanding CNN processing performance with excellent power efficiency.