AI-focused MCUs embed neural processor
Aimed at AI/ML applications, Renesas’ RA8P1 MCUs leverage an Arm Ethos-U55 neural processing unit (NPU) delivering 256 GOPS at 500 MHz. The 32-bit devices also integrate dual CPU cores—a 1-GHz Arm Cortex-M85 and a 250-MHz Cortex-M33—that together achieve over 7300 CoreMark points.
The NPU supports most commonly used neural networks, including DS-CNN, ResNet, Mobilenet, and TinyYolo. Depending on the neural network used, the Ethos-U55 provides up to 35× more inferences per second than the Cortex-M85 processor on its own.
RA8P1 microcontrollers provide up to 2 MB of SRAM and 1 MB of MRAM, which offers faster write speeds and higher endurance than flash memory. System-in-package options include 4 MB or 8 MB of external flash memory for more demanding AI tasks.
Dedicated peripherals and advanced security features support voice and vision AI, as well as real-time analytics. For vision AI, a 16-bit camera engine (CEU) handles image sensors up to 5 megapixels, while a separate two-lane MIPI CSI-2 interface provides a low pin-count connection at up to 720 Mbps per lane. Audio interfaces including I²S and PDM enable microphone input for voice AI. To protect edge AI and IoT systems, the devices integrate cryptographic IP, enforce immutable storage, and monitor for physical tampering.
The RA8P1 MCUs are available now in 224-pin and 289-pin BGA packages.
The post AI-focused MCUs embed neural processor appeared first on EDN.
