Renesas Electronics announced the new 64-bit RZ/G3E microprocessor (MPU), designed for high-performance HMI systems requiring AI acceleration and edge computing. Based on the quad-core Arm Cortex-A55 processor and the Ethos-U55 Neural Processing Unit (NPU), the new MPU efficiently runs AI applications such as image classification, object recognition, speech recognition, and anomaly detection while reducing CPU load.
Efficient Edge Computing and HMI Functionality
The RZ/G3E is a general-purpose product optimized for high-performance human-machine interface (HMI) applications. It integrates a quad-core Arm Cortex-A55 running at up to 1.8GHz and a neural network processing unit (NPU). This delivers high-performance computing and AI inference capabilities, enabling faster and more efficient local processing. With its full-HD graphics capabilities and high-speed connectivity, this MPU is primarily targeted at HMI systems in industrial and consumer sectors, including factory equipment, medical monitors, retail terminals, and building automation systems.
HMIs are an integral part of modern automation systems and intelligent devices. As a bridge between operators and machines, they enable users to effectively monitor, control, and understand complex equipment and processes, significantly improving efficiency, safety, and operational convenience.
Designed specifically for HMI applications, this product delivers smooth Full HD (1920x1080) video playback at 60fps on two independent displays. Output interfaces include LVDS (dual-link), MIPI-DSI, and parallel RGB. Additionally, the RZ/G3E features an MPI-CSI camera interface for video input sensing applications.
“The RZ/G3E builds on the proven performance of the RZ/G series by adding an NPU to support AI processing,” said Daryl Khoo, Vice President of Embedded Processing at Renesas. “By leveraging the same Ethos™-U55 NPU as our recently released RA8P1 microcontroller, we’ve enabled next-generation development with the range of high-speed communication interfaces required for RZ/G3E-equipped edge devices, including PCI Express 3.0 (2 lanes) for up to 8Gbps data transfer, USB 3.2 Gen2 for blazing-fast 10Gbps data transfer, and dual-lane Gigabit Ethernet for seamless connectivity to cloud services, storage devices, and 5G modules.”
Low-Power Standby and Fast Linux Resume
Beginning with the third-generation RZ/G3E, the RZ/G Series has included advanced power management features that significantly reduce standby power consumption. The RZ/G3E maintains sub-CPU operation and peripheral functions while consuming as little as approximately 50mW. Deep standby mode consumes approximately 1mW. It supports DDR self-refresh mode to retain memory data, enabling fast wakeup from deep standby mode to run Linux applications.
Renesas offers a verified Linux Software Package (VLP) based on the reliable Civil Infrastructure Platform core, with over 10 years of maintenance support. For users requiring the latest version, Renesas offers Linux BSP Plus, which supports the latest LTS Linux kernel and Yocto. Canonical's Ubuntu and Debian open source operating systems are also available for server or desktop Linux environments.
Key Features of the RZ/G3E
CPU: Quad-core Cortex®-A55 (up to 1.8GHz), Cortex®-M33
NPU: Ethos™-U55 (512 GOPS)
HMI: Dual Full HD outputs, MIPI-DSI/Dual-link LVDS/Parallel RGB, 3D graphics processing, H.264/H.265 codecs
Memory Interface: 32-bit LPDDR4/LPDDR4X (with ECC)
5G Communication: PCIe 3.0 (2 lanes), USB 3.2 Gen2, USB 2.0 x2, Gigabit Ethernet x2, CAN-FD
Operating Temperature: -40°C to 125°C
Packaging Options: 15mm² 529-pin FCBGA, 21mm² 625-pin FCBGA
Product Lifecycle: 15-year supply guarantee under the Product Lifecycle Plan (PLP)
Source from Internet
The RZ/G3E is available now, along with an evaluation kit including a SMARC v2.1.1 module and carrier board. Renesas also announced a System-on-Module (SoM) solution based on the RZ/G3E: the Renesas High-Performance Edge Computing SoM. Renesas' ecosystem partners offer a wide range of SoM solutions, including Tria's SMARC modules, ARIES Embedded's OSM (Size-M), and MXT's OSM (Size-L).