SoC engineered for edge deep learning applications

March 30, 2021
The R-Car V3H includes dedicated hardware acceleration and an image recognition engine.

The R-Car V3H System-on-Chip (SoC), designed for front-facing cameras for autonomous vehicles and robotics applications, features four 1.0 GhZ Arm Cortex-A53 MPCore processors, a dual lockstep 800 MHz Arm Cortex-R7 core, and a memory controller for LPDDR4-3200 with 32 bits x 1 channel.

The SoC also features a dual image signal processor, 4 lanes x 2 channels MIPI-CS12 and 2 channel digital video inputs, and 4 lanes x 1 channel LVDS and 1 channel digital video output. Available interfaces include GigE, AVB Ethernet, and PCIe.

Dedicated hardware acceleration for convolutional neural networks (CNNs), dense optical flow, stereo vision, and object classification, along with the IMP-X5-V3H image recognition engine, make the R-Car V3H appropriate for edge applications running deep learning software.


To Learn More:

Contact: Renesas Electronics
Headquarters: Koto City, Tokyo, Japan
Product: R-Car V3H Soc
Key Features: Dedicated hardware acceleration for deep learning software, GigE, AVB Ethernet and PCIe interfaces, IMP-X5-V3H image recognition engine, ARM processors.

What Renesas Electronics says:
View more information on the R-Car V3H SoC.

Voice Your Opinion

To join the conversation, and become an exclusive member of Vision Systems Design, create an account today!