Home > Products > Preferred toolchain selected for new neural network-based AI SoCs

Design

Preferred toolchain selected for new neural network-based AI SoCs

25th January 2019

IAR Systems

Alex Lynn

0 0

The supplier of software tools and services for embedded development, IAR Systems, has announced that it has been selected by Eta Compute as the preferred tools vendor for its new machine-learning SoCs. Using IAR Embedded Workbench for Arm with Eta Compute SoCs, developers should be able to leverage the power-efficient features of the SoC, boosting development of AI-based IoT applications.

Eta Compute is part of the Arm DesignStart program, which enables the creation of custom SoCs with low cost and fast access, and provides a verified subsystem for a wide range of applications, including IoT, gateways, sensor, control and mixed signal SoCs, as well as design services and support.

At Arm TechCon in October 2018, Eta Compute won the Design Innovation of the Year and the Best Use of Advanced Technologies awards. By implementing the popular and proven Arm Cortex-M3 architecture with improved power efficiency, Eta Compute opened up opportunities for applying more comprehensive AI processing on edge devices.

“IAR Systems’ tools have a demonstrated capability of generating the smallest, most efficient code, while maintaining application integrity,” said Paul Washkewicz, Vice President and Co-Founder of Eta Compute. “IAR Embedded Workbench is fast and lightweight, resulting in an IDE that aids in development. Additionally, their quick, knowledgeable, and professional support teams help accelerate the time it takes to move from prototype to production.”

Robert DeOliveira, Director Strategic Sales, IAR Systems, added: “Eta Compute is an exciting startup that has accomplished game changing technology in the three years since its start. They are actively exploring new application areas for their technology. Their decision to standardise development on IAR Embedded Workbench underscores the importance of having sophisticated, smart tools to overcome challenges and complexities in next-generation embedded applications.”

IAR Embedded Workbench is a C/C++ compiler and debugger toolchain for developing Arm Cortex-based applications. Powerful add-ons and integrations, including easy-to-use debugging probes and integrated code analysis tools, add additional capabilities.

Eta Compute’s DIAL Technology is a low power embedded platform that delivers significantly more machine intelligence to energy-constrained products. For a given power source, DIAL delivers the most compute power enabling local intelligence on the device, which increases accuracy and speed of decisions. Processing at the network edge also reduces power hungry RF transmissions extending battery life while improving security and privacy.