Renesas Electronics Corporation, a provider of advanced semiconductor solutions, and Fixstars Corporation, a global provider of multi-core CPU/GPU/FPGA acceleration technology, announced the joint development of a suite of tools that enables rapid software optimization and simulation for autonomous driving (AD) and advanced driver assistance systems (ADAS). The technology was designed specifically for Renesas' R-Car system-on-chip (SoC) devices.
These tools allow you to quickly develop network models with highly accurate object recognition from the early stage of software development that take advantage of R-Car's performance. This reduces post-development rework and therefore helps shorten development cycles.
“Renesas continues to create integrated development environments that enable customers to adopt a 'software first' approach,” said Hirofumi Kawaguchi, vice president of Renesas' Automotive Software Development Division. “By supporting the development of deep learning models tailored to R-Car, we help our customers build AD and ADAS solutions while reducing time to market and development costs.”
Today's AD and ADAS applications use deep learning to achieve highly accurate object recognition. Deep learning inference processing requires large amounts of data calculations and memory capacity. Models and programs executable in automotive applications must be optimized for an automotive SoC, since real-time processing with limited arithmetic units and memory resources can be a challenging task.
Furthermore, the process from software evaluation to verification must be accelerated and updates must be applied repeatedly to improve accuracy and performance. Renesas and Fixstars have developed the following tools designed to meet these needs.
“GENESIS for R-Car, which is a cloud-based evaluation environment we built together with Renesas, allows engineers to evaluate and select devices early in development cycles and has already been used by many customers,” said Satoshi Miki, CEO of Fixstars. “We will continue to develop new technologies to accelerate machine learning operations (MLOps) that can be used to maintain the latest versions of software in automotive applications.”
1. R-Car Neural Architecture Search (NAS) Tool — Generates deep learning network models that efficiently use the CNN (convolutional neural network) accelerator, DSP, and memory on the R-Car device. This allows engineers to quickly develop lightweight network models that achieve highly accurate object recognition and fast processing time, even without in-depth knowledge or experience with the R-Car architecture.
2. R-Car DNN Compiler to Compile Network – Converts optimized network models into programs that can take full advantage of R-Car's performance potential. It converts network models into programs that can be quickly executed on CNN IP and also performs memory optimization to enable high-speed, limited-capacity SRAM to maximize its performance.
3. R-Car DNN Simulator – can be used to quickly verify the operation of programs on a PC rather than on the actual R-Car chip. Using this tool, developers can generate the same operational results that would be produced by R-Car. If the recognition accuracy of inference processing is affected during the process of lightening models and optimizing programs, engineers can provide immediate feedback for model development, thereby shortening development cycles.
Renesas and Fixstars will continue to develop software for deep learning with the joint Automotive SW Platform Lab and build operating environments that maintain and improve recognition accuracy and performance by continually updating network models.
