Akida Pico: groundbreaking ultra-low power AI coprocessor

BrainChip, the pioneer of ultra-low power brain-inspired AI technology, has once again pushed the boundaries of AI co-processing with the launch of Akida Pico. As the world’s first commercial manufacturer of ultra-low power event-based AI chips, BrainChip has just delivered Akida Pico, a lowest-power accelerated coprocessor designed specifically for the next wave of smart, portable devices. This innovation has the potential to revolutionize various sectors including healthcare, IoT, consumer electronics and defense by providing an energy-efficient solution for edge AI applications.

With power consumption of less than one milliwatt, Akida Pico offers a fully digital, compact and secure architecture that is tailored to applications requiring neural network acceleration in often battery-powered devices. Whether it’s voice wake-up detection, automatic doorbells, or advanced wearable AI, the Akida Pico delivers exceptionally energy-efficient performance in a compact form factor.

BrainChip Akida Pico

This advancement relies on BrainChip’s Akida 2 event-based computing platform, which enables battery-powered systems to operate continuously with minimal energy consumption. By filtering false alarms, Akida Pico helps save energy until significant events are detected, making it ideal for applications where continuous monitoring is required.

Energy-efficient AI processing for edge devices

The Akida Pico coprocessor is specifically designed to accelerate task-limited neural network models, ensuring optimal energy efficiency. It is the embodiment of both the digital and event-driven architecture that BrainChip is pioneering in supporting edge devices where energy efficiency is paramount.

A key feature of the Akida Pico is its ability to maintain functionality with power consumption of less than one milliwatt, allowing it to operate on battery power without frequent recharging. This makes it perfect for devices that require consistent performance but cannot afford high energy consumption. For example, the healthcare and IoT industries rely on sensors and monitoring systems that need to be both responsive and energy efficient, and the Akida Pico seamlessly meets these requirements.

In practice, Akida Pico is expected to be used in voice assistants, speech noise reduction, presence detection and many others. BrainChip’s Akida Pico can act as a co-processor that wakes up larger system processors when needed, while filtering out extraneous signals that could otherwise extend battery life. This “wake up” feature is essential for devices such as wearables that need to run continuously but only occasionally require intensive processing.

Developer-friendly and flexible thanks to MetaTF

One of the strongest aspects of Akida Pico is its compatibility with standard development environments. BrainChip’s MetaTF software flow enables developers to easily compile and optimize temporally aware neural networks (TENNs) without having to adopt a new machine language framework. Seamless integration with widely used tools such as TensorFlow/Keras and PyTorch ensures developers can continue working in familiar environments while taking full advantage of the power and performance that Akida Pico offers.

The MetaTF framework simplifies the process for both novice and experienced developers by making it easier to create, train, and deploy AI models designed for specific edge applications. With this ease of use, BrainChip’s Akida Pico opens the door to a wider range of AI solutions, enabling developers with varying levels of expertise to contribute to innovation in wearable technology, home automation, and more.

MetaTF’s flexibility also means developers can build AI models that suit the specifically event-driven nature of Akida Pico, optimizing their applications for low power consumption and real-time responsiveness. The result is a highly adaptable AI system that can serve a variety of industries without compromising performance.

Impact on many industries

Akida Pico has the potential to make a significant impact in several key sectors where energy efficiency and continuous monitoring are crucial. In healthcare, for example, mobile devices that need to track vital signs or detect abnormal conditions such as arrhythmias can benefit from Akida Pico’s low power consumption and real-time processing capabilities.

Similarly, for IoT devices such as smart home systems and security solutions, the Akida Pico’s “wake up” feature will be invaluable. These devices often remain in a sleep state until triggered by specific events, and Akida Pico’s ability to filter false alarms ensures that these systems remain efficient by activating the full processor only when necessary.

In the defense field, continuous monitoring systems, which require robust but energy-efficient artificial intelligence, are also potential beneficiaries. By offering a co-processor that can handle tasks such as audio recognition, object tracking or anomaly detection with minimal power consumption, BrainChip’s Akida Pico positions itself as an invaluable tool in the field of cutting-edge defense technologies.

Akida Pico also represents a breakthrough in consumer electronics, especially in devices integrating voice assistants or artificial intelligence-based interfaces. As demand for smarter, more intuitive gadgets increases, the low latency and energy efficiency of Akida Pico will play a key role in meeting consumer expectations while extending battery life.

BrainChip’s Akida Pico is a new innovation in the world of AI co-processing. By offering an ultra-low power solution that is flexible, developer-friendly and applicable across a variety of industries, Akida Pico enables a new generation of smart devices. This is more than just advancements in AI technology; is a step towards creating smarter, more sustainable and more efficient systems that can operate at the edge. With Akida Pico, the future of energy-saving devices powered by artificial intelligence has arrived. Here is a selection of other articles from our extensive content library that may interest you about AI processors:

Filed Under: Equipment, Top news

The latest gadget offers for geeks

Disclosure: Some of our articles contain affiliate links. If you purchase something through one of these links, Geeky Gadgets may receive an affiliate commission. Learn about our Disclosure Policy.