Power Efficient MPUs With High-Precision AI Accelerator

360
 

Both AI inference and image processing functions are handled, eliminating external ISP requirements and implementing cost-efficient Vision AI

Renesas Electronics Corporation announced the expansion of its RZ/V Series of microprocessors (MPUs) with the new RZ/V2L MPUs designed for entry-level AI-enabled applications. As part of the RZ/V Series, the new MPUs incorporate the Dynamically Reconfigurable Processor (DRP-AI) – an AI accelerator that makes embedded AI easier and more power-efficient. The new RZ/V2L shares several features with its predecessor, the RZ/V2M, such as high-precision AI inference capabilities combined with top-class power efficiency. It also includes optimisations such as tailoring the DRP-AI operating frequency and memory interface for an entry-level MPU.

The DRP-AI provides both real-time AI inference and image processing functions with the capabilities essential for camera support such as colour correction and noise reduction, enabling customers to develop AI-based vision applications, such as point-of-sale (POS) terminals and robot vacuum cleaners, without requiring an external image signal processor (ISP). Additionally, the RZ/V2L’s excellent power efficiency eliminates the need for heat dissipation measures such as heat sinks or cooling fans. Now AI can be implemented cost-efficiently not only in surveillance cameras and industrial equipment but also in a wide range of applications including home appliances and consumer electronics.

The RZ/V2L is also package and pin-compatible with the existing RZ/G2L general-purpose MPUs. This allows RZ/G2L users to easily upgrade to the RZ/V2L for additional AI functions without needing to modify the system configuration, keeping migration costs low.

“One of the most important trends in the electronics industry today is the incorporation of AI into embedded devices – especially AI for interpreting sensor data, such as images,” said Jeff Bier, General Chairman of the Embedded Vision Summit and Founder of the Edge AI and Vision Alliance. “Innovative processors like the RZ/V2L are key for enabling perceptual AI to be incorporated into a wide range of cost and power-constrained IoT devices.”

“Our new entry-level AI accelerator products provide customers with the flexibility to select the most suitable product from a wide range of MPU offerings,” said Hiroto Nitta, Senior Vice President and Head of SoC Business in the IoT and Infrastructure Business Unit at Renesas. “These AI-optimised solutions will expand the use of embedded AI to a wider range of applications such as home appliances and consumer electronics, as customers can now implement AI capability with lower system costs.”

As part of the RZ/V2L development environment, Renesas offers a complimentary DRP-AI Translator, a tool that automatically converts AI models into an executable format. The input format is the industry-standard Open Neural Network Exchange (ONNX). Developers can leverage the DRP-AI while using the tools they are accustomed to, allowing them to immediately start using the RZ/V2L to evaluate AI models based on proven learning data.

Key Features of the RZ/V2L MPUs

  • 64-bit Arm Cortex-A55 (1.2 GHz, dual or single-core) and Cortex-M33
  • DRP-AI (1 TOPS/W class) AI accelerator capable of running the Tiny YOLOv2 program at 28 frames per second (fps)
  • Simple ISP functions required for machine vision are provided in the DRP library (supports up to full HD)
  • 16-bit, single-channel DDR memory interface
  • 3D graphics functions (Arm Mali-G31 GPU)
  • Video codec (H.264)
  • CMOS sensor interfaces (MIPI-CSI and Parallel) for camera input
  • Display interfaces (MIPI-DSI and Parallel)
  • Memory with error checking and correction (ECC)
  • Verified Linux Package (VLP) based on Civil Infrastructure Platform Linux, an industrial-grade Linux implementation, available
  • Available in 15 mm square or 21 mm square BGA packages that are pin-compatible with the RZ/G2L

RZ/V2L evaluation boards are provided as “Smart Mobility ARChitecture (SMARC) System-on-Module (SoM) Solution” to enable quick validation for different applications. SMARC Solution is also available for existing RZ/G2 MPUs. These reference designs provide an optimised circuit diagram and board layout including a power supply circuit and timing tree. Furthermore, a power management IC (PMIC) optimised for the RZ/V2L is currently in development. Solutions incorporating the RZ/V2L and the new PMIC are expected to become available in the second half of 2021. 

Sample shipments of the RZ/V2L start today, and mass production is scheduled to begin in December 2021.


SHARE YOUR THOUGHTS & COMMENTS

Please enter your comment!
Please enter your name here