Renesas Intros 64-bit MPU Aimed at AI-Centric High-Performance HMI Designs

Renesas has launched a premier MCU tailored for high-performance HMI applications and edge computing.

Industrial and consumer sectors require HMI systems that support Full HD graphics and high-speed connectivity, which is what this MPU is designed for.

Additionally, for faster and more efficient local processing, it enables high-performance edge computing with AI inference.

The Reneas RZ/G3E MPU offers features specifically for high performance HMI applications.
 

Outstanding Features for HMI Systems

The RZ/G3E combines a quad-core Arm Cortex-A55 running at up to 1.8 GHz, a Cortex-M33 core to enable low power operation, and the Ethos-U55 NPU with a performance of 512 GOPS for AI acceleration.

Focused on HMI applications, it includes dual Full HD (1920 x 1080) video output at 60 fps on independent displays, LVDS (dual link), MIPI-DSI and parallel RGB output interfaces, 3D graphics engine, and H.264/H.265 video codec. Meanwhile, a MIPI-CSI camera interface is provided for developing video input and detection applications.

For fast wake-up from deep standby to run applications, 32-bit LPDDR4/LPDDR4X with ECC RAM is built in, which maintains memory data through the DDR auto-refresh mode.

RZ/G3E block diagram. (Click on image to enlarge)

On-Chip Technology for AI Processing

For AI application development, the RZ/G3E is powered by a quad-core Arm Cortex-A55 processor, a Cortex-M33 core, and the Ethos-U55 NPU.

By minimizing CPU overhead, this processor enables the execution of high-performance AI applications, including object and speech recognition, image classification, and anomaly detection. With powerful AI capabilities, it's possible to design advanced next-generation HMI projects in fields such as vision, voice, and real-time analytics.
 

Significant Set of Interfaces

To get the most out of edge devices, the new MPU features a set of high-speed communication interfaces, including:

• Two 8 Gbps PCI Express 3.0 lanes
• Dual-channel Gigabit Ethernet for efficient connectivity with cloud services, 5G modules, and storage
• USB 3.2 Gen2 at 10 Gbps for fast data transfers
• 2-channel USB 2.0 and SD interface
• CAN interface
• LPDDR4/LPDDR4X memory interface

More information can be found in the RZ/G3E data sheet.

Application Example

A digital otoscope, which allows for accurate, high-resolution, and minimally invasive visualization of the middle ear canal, is crucial for accurate diagnosis and effective treatment.

The design incorporates a Full HD video display with H.264 and H.265 support that delivers crystal-clear, real-time images to identify blockages or signs of infection in the ear canal, powered by a dual or quad Arm Cortex-A55 CPU.

To identify obstructions and symptoms of infection, AI-based inference with an Arm Ethos-U55 NPU can be used, enabling real-time analysis and improving diagnostic support for clinicians.

For documentation and medical record-keeping tasks, it allows image capture and video recording via built-in SD cards. For mobile exams, it enables wireless video transmission to smartphones or tablets thanks to the built-in low-power Wi-Fi/Bluetooth module.

Block diagram of digital otoscope using the RZ/G3E MPU. (Click on image to enlarge)

Beyond the Limits

Clearly the new general-purpose RZ/G3E MPU addresses demands for high-performance HMI designs, leveraging the device’s integrated NPU for edge computing. 

This device enables high-speed connectivity and advanced graphics across diverse applications, paving the way for future innovations in HMI systems requiring AI acceleration.

All Images used courtesy of Renesas.