Synopsys has formed a collaboration with Kyocera Document Solutions, enabling Kyocera to achieve first-pass silicon success for its new multifunction product (MFP) SoC using Synopsys DesignWare ARC EV6x Embedded Vision Processor IP with convolutional neural network (CNN) engine and ARC MetaWare EV Development Toolkit.
By using Synopsys’ programmable ARC EV Processor, Kyocera successfully integrated high-performance artificial intelligence (AI) processing capabilities such as super-resolution, with the flexibility to support future AI models.
“Implementing advanced AI functionality into our MFP SoC required high-performance, low-power processor IP with a high-quality toolchain, allowing us to find and test AI algorithms while developing the SoC in parallel,” said Michihiro Okada, General Manager, Software Development Division at Kyocera Document Solutions Inc. “Only Synopsys’ DesignWare ARC EV Processor IP and mature MetaWare EV Toolkit met our extensibility, performance, and area requirements.”
“As new artificial intelligence applications incorporate the latest high-performance neural networks, designers require superior hardware technology and software tools to deliver SoCs on schedule,” said John Koeter, Senior Vice President of Marketing and Strategy for IP at Synopsys. “With our AI-optimized IP solutions that enable fast, accurate vision processing, companies like Kyocera can build competitive products faster and with significantly less risk.”
Also, Kyocera deployed Synopsys’ HAPS FPGA-based prototyping system to accelerate ARC EV software development, SoC integration, and system validation.
The DesignWare ARC EV6x Processor family integrates scalar, vector DSP, and CNN processing units for highly accurate and fast vision processing. The EV6x Processors support popular convolutional neural networks including SR-GAN, MobileNET, Yolo, GoogLeNet, SSD, and ResNet.
To simplify software application development, the EV6x processors are supported by the DesignWare ARC MetaWare EV Development Toolkit, a comprehensive suite of tools and software.
The toolkit includes the MetaWare NN Compiler that analyzes neural networks trained using popular frameworks like Tensorflow or Pytorch and automatically generates the executable for the programmable CNN engine.