Deploying adaptive AI in distributed water plants - Barbara Accoina

BrainChip aims to amplify edge AI with second-gen Akida platform

BrainChip aims to amplify edge AI with second-gen Akida platform

BrainChip, a neuromorphic computing device provider, has unveiled the early access release of its second-generation Akida IP solution. The company developed this IP solution to cater to various applications across the smart home, smart city, industrial and automotive sectors.

“This is a significant step in BrainChip’s vision to bring unprecedented AI processing power to edge devices, untethered from the cloud,” states Sean Hehir, the CEO of BrainChip.

According to BrainChip, the second-generation Akida platform enables energy-efficient processing of complex neural network models on edge devices. With support for 8-bit weights, activations and long-range skip connections, it expands the capabilities of models accelerated in Akida’s hardware. Brainchip says this technology effectively caters to the rising demand for advanced and efficient computing at the edge.

According to company executives, implementing Temporal Event Based Neural Nets (TENNs) transforms advanced sequential processing for multi-dimensional streaming and time-series data. This innovation aims to significantly reduce model size while enhancing performance and efficiency.

Combining hardware acceleration of Vision Transformers (ViT) models with edge devices further unlocks the potential for game-changing processing of vision, video, audio and other applications.

According to Jean-Luc Chatelain, the managing director of Verax Capital Advisors, generative AI and LLMs at the edge play a crucial role in intelligent situational awareness across various industries such as manufacturing, healthcare and defense.

“BrainChip TENNs support Vision Transformers built on the foundation of neuromorphic principles and can deliver compelling solutions in ultra-low power, small form factor devices at the edge, without compromising accuracy,” he explains.

Further, the MetaTF software, now in its second generation, empowers developers to assess Akiba’s capabilities, enhance designs and customize System-on-chip (SoC) and software solutions. With support for both TensorFlow and ONNX, MetaTF offers compatibility across multiple frameworks, including PyTorch, company executives state.

Zach Shelby, CEO of Edge Impulse, emphasizes that multimodal edge AI is an irreversible trend that places greater demands on the intelligent compute capabilities of edge devices. Further, he emphasizes that the second generation Akida prioritizes performance, efficiency, accuracy and reliability to accelerate the transition.

BrainChip adds that Akida processors drive the growth of edge AI devices in various environments like industrial, home, automotive and IoT. With its digital, customizable, event-based neural processing solution, use cases include intelligent sensing, medical monitoring and video-object detection.

BrainChip recently formed a strategic partnership with VVDN Technologies to develop the Edge Box. This edge device processes data and performs computations at the network edge. The companies say that this hardware platform utilizes neuromorphic technology inspired by the human brain, resulting in improved power efficiency and performance for edge applications.

Article Topics

 |   |   | 

Comments

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Featured Edge Computing Company

Edge Ecosystem Videos

Machine learning at the Edge

“Barbara

Latest News