Eta Compute’s integrated AI vision card lasts “years” on a single battery



Article by: Nitin Dahad

Eta Compute has launched an integrated AI-integrated vision card that it says can enable vision applications that can last for years on a single battery.

Eta Compute has launched an artificial intelligence (AI) integrated vision card that the company says can enable vision applications that can last for years on a single battery.

Eta Compute - Eagle with Sparkfun
The board includes three sensors, a low-power Himax camera and an expansion slot. (Source: Eta Compute)

The small form factor (1.5 “x1.5”) of its AI ECM3532 vision board, built-in battery, and low-power Internet of Things (IoT) connectivity and Bluetooth Low-Energy make it suitable for prototyping, field testing and deployment of the integrated vision to AI. applications. The board includes three sensors (ambient light, microphone, accelerometer / gyroscope), a low power Himax HM0360 camera and an expansion slot.

The company said its ultra-low-power operation removes barriers with traditional connected solutions or boards that have extremely limited battery life and high power consumption. The AI ​​vision card is the second in its family of cards, modules and systems designed by Eta Compute.

[Webinar]Everyone and everything, connected

The new board is supported by Edge Impulse’s machine learning (ML) development platform for rapid neural network development, making the design of energy-efficient vision endpoints transparent. The companies have collaborated to integrate Eta Compute’s TENSAI Flow software, optimizing design flow for increased efficiency in the on-board AI design of next-generation smart devices. This software enables developers to quickly verify feasibility and proof of concept, and enables seamless design from concept to firmware, for building ML applications in IoT and low power edge devices. TENSAI Flow includes neural network compiler, neural network zoo and middleware including FreeRTOS, hardware abstraction level (HAL) and frameworks for sensors, as well as IoT / cloud activation.

Eta Compute Tensai Platform
Eta Compute TENSAI platform (Source: Eta Compute)

Compared to the direct implementation on a competitor device of the same CIFAR10 neural network, the TENSAI Neural Network Compiler on TENSAI SoC (System on Chip) improves inference energy by a factor of 54. Using the CIFAR10 neural network from the TENSAI Neural Network Zoo and TENSAI Neural Network Compiler further improves inference energy, bringing this figure to a factor of 200x.

By interfacing with Edge Impulse, TENSAI Flow enables developers to securely acquire and store training data so that customers practice once and have actual models for future development. The software automatically optimizes TensorFlow Lite AI models for Eta Compute’s TENSAI SoC, delivering high optimization and energy efficiency. With TENSAI Flow, TENSAI SoC can load AI models that include sensor interfaces seamlessly. TENSAI Flow provides the foundation for automatically provisioning and connecting devices to the cloud and upgrading firmware over the air based on new models or data.

Eta Compute power consumption comparison
The TENSAI platform enables AI performance in the order of 1 milliwatt, which is extremely low compared to other solutions, especially for image processing. (Source: Eta Compute)

Explaining the importance of low power integrated vision capability, Jeff Bier, Founder of Edge AI and Vision Alliance, said, “With computer vision, devices can understand the world around them, which enables them to be more efficient, safer, easier to use and more autonomous. But vision algorithms are very computationally intensive, and the power consumption required to provide the necessary processing performance has made these capabilities impractical for many potential applications. We applaud Eta Compute’s innovation and collaboration with other Edge AI and Vision Alliance companies, which make vision achievable in many new power-sensitive applications.

Edge Impulse CEO and Co-Founder Zach Shelby commented, “With Edge Impulse, Eta Compute developers can design, test and deploy fast, embedded applications across a multitude of workloads, from detection to ‘objects to classification and actual counting, through human, animal and machine spectra. Ted Tewksbury, CEO of Eta Compute, added: “For the first time, they can leverage an integrated board complemented by Edge Impulse’s machine learning development platform to deploy vision applications that can have the power to transform people’s lives and work.

This article originally appeared on Embedded.

Nitin Dahad is editor of and correspondent for EE Times and EE Times Europe. Since starting his career in the electronics industry in 1985, he has held many different roles: from engineer to journalist, and from entrepreneur to startup mentor and government advisor. He was part of the start-up team that launched and publicized 32-bit microprocessor company ARC International in the United States in the late 1990s, and co-founder of The Chilli, which influenced a great deal part of the tech startup scene in the early 2000s. He has also worked with many big names including National Semiconductor, GEC Plessey Semiconductors, Dialog Semiconductor and Marconi Instruments.

Congratulations to the winners of the 1st round! Round 2 starts now!

    Prize draw for 2021 members



Leave A Reply