Microprocessor delivers high-speed e-AI imaging at low power consumption

By Alex Pluemer, contributing writer

Renesas Electronics Corp. has unveiled a new microprocessor for embedded artificial intelligence (e-AI) image processing for applications such as smart appliances, service robots, and industrial machinery. The RZ/A2M microprocessor implements Renesas’ proprietary Dynamically Reconfigurable Processor (DRP) to deliver real-time image processing while significantly reducing power consumption compared with its predecessors.

The RZ/A2M makes external DRAM unnecessary by being outfitted with large-capacity (4-MB) on-chip RAM and supports the MIPI camera interface commonly found in smartphones and other mobile devices. The RZ/A2M also boasts increased network functionality through the incorporation of two-channel Ethernet support and greater network security by implementing an on-chip hardware encryption accelerator.

The RZ/A2M with DRP enables image-based AI functionality involving large amounts of data and more powerful processing performance than achievable with conventional waveform measurement and analysis. By reducing the amount of power consumed during real-time image processing, the RZ/A2M will allow battery-powered devices to allocate power to perform other operations such as biometric authentication via fingerprint or retinal scans.

“Because real-time image processing can be accomplished while consuming very little power, battery-powered devices can perform tasks such as real-time image recognition based on camera input, biometric authentication using fingerprints or iris scans, and high-speed scanning by handheld scanners,” said Renesas. “This solves several issues associated with cloud-based approaches, such as the difficulty of achieving real-time performance, assuring privacy, and maintaining security.”

Samples of the RZ/A2M are currently available from Renesas, with mass production scheduled to begin in Q1 2019. Monthly volume production for all RZ/A2M versions is expected to reach 400,000 units by 2021. The MPUs are accompanied by a DRP image-processing library, reference software, and a development board to assist developers in evaluating HMI function and image-processing functionality and performance.