Automotive SoCs target ADAS

Renesas Electronics Corp. has expanded its R-Car family of automotive system-on-chips (SoCs) for entry-level advanced-driver assistance systems (ADAS). The new devices includes the new R-Car V4M series and the expansion of the existing R-Car V4H series of automotive SoCs for high-volume L2 and L2+ ADAS. These devices offer AI processing capability, fast CPU performance and power-optimized features that make them suited for cost-sensitive ADAS applications such as front smart camera systems, surround-view systems, automatic parking and driver monitoring systems.

(Source: Renesas Electronics Corp.)

The R-Car V4M and R-Car V4H devices offer advanced AI technology, performance and power efficiency to meet a variety of ADAS application requirements. The new additions give customers greater flexibility to select the best SoC that meets their requirements.

The new devices offer pin compatibility within the same series and software compatibility with existing R-Car products. This allows OEMs and Tier-1 suppliers to reuse existing software and reduce engineering costs, Renesas said.

Based on the same advanced 7-nm manufacturing process technology as the R-Car V4H series, the R-Car V4M series offers deep learning performance of up to 17 tera operations per second (TOPS) and high-speed image processing and precise object recognition using on-board cameras, radar and LiDAR.

Both the R-Car V4M and R-Car V4H series feature up to four Arm Cortex-A76 cores, delivering up to 81K DMIPS performance for application processing. It also comes with three Arm Cortex-R52 lockstep cores with up to 25K DMIPS performance for real-time operation.

The new devices are power efficient, providing 9 TOPS/W performance, thanks to their highly integrated design and manufacturing process technology. In a typical full-feature smart camera with an 8-megapixel sensor, the R-Car V4M consumes around 5 watts of power, which is 50 percent less than similar devices on the market, according to Renesas.

The devices integrate dedicated deep learning and computer vision IP; an image signal processor with parallel processing for machine and human vision; an image renderer for fisheye distortion correction or other mathematical operations; and a 3D graphic processor unit. Other features include fast boot (<1 second) for a camera display, expandable memory size AUTOSAR software partition and several automotive interfaces, including CAN, Ethernet AVB, TSN and FlexRay.

The R-Car V4M series and R-Car V4H series can leverage Renesas’ software-defined vehicle (SDV) development environment. The R-Car Open Access (RoX) SDV platform integrates all essential hardware, operating systems (OS), software and tools needed to develop SDVs. With the development platform, OEMs and Tier 1 suppliers can design a range of scalable compute solutions for ADAS, IVI, gateway and cross-domain fusion systems as well as body control, domain and zone control systems The platform comes with a software development kit (SDK) and the AI Workbench, which enables developers to validate and optimize their models and test their applications in the cloud.

Renesas is sampling the R-Car V4M and R-Car V4H devices to leading automotive manufacturers now. Mass production is scheduled in the first quarter of 2026. Development is underway for the 5^th Generation R-Car SoCs for applications in ADAS, cockpit, gateway and infotainment segments.