TL;DR:
- BrainChip unveils a second-generation Akida platform for Edge AI.
- Akida 2.0 promises an “order of magnitude” improvement in compute density.
- It supports Temporal Event-Based Neural Network (TENN) acceleration and vision transformer hardware.
- Akida 2.0 includes three product classes: Akida-E, Akida-S, and Akida-P.
- Early access program initiated for Akida 2.0 intellectual property.
- Signifies a significant shift towards Multimodal Edge AI.
- TENNs offer a substantial reduction in model size and computational demands.
- The availability of hardware development kits is yet to be confirmed.
Main AI News:
In a world driven by an insatiable hunger for artificial intelligence (AI) capabilities, BrainChip, a trailblazer in neural network processors, has taken a momentous step toward empowering edge devices with unparalleled processing prowess. The company’s latest revelation, the second-generation Akida platform, signifies a revolutionary leap in the realm of Edge AI, offering the potential to liberate devices from their reliance on the cloud.
The initial unveiling of BrainChip’s original Akida neuromorphic processing technology, showcased at the Linley Fall Processor Conference in 2019, set the stage for a journey that culminated in the availability of development kits for the general public in 2021. Fast forward to March 2023, and the announcement of Akida 2.0, a refinement promising support for Temporal Event-Based Neural Network (TENN) acceleration, as well as optional vision transformer hardware. This enhancement not only amplifies the platform’s capabilities but also alleviates the computational burden on the host processor. BrainChip has categorized Akida 2.0 into three distinct product classes: Akida-E, emphasizing energy efficiency; Akida-S, designed for seamless integration into microcontroller units and systems-on-chips; and Akida-P, a high-performance range complemented by optional vision transformer acceleration.
Now, BrainChip has initiated an “early access” program, providing access to the Akida 2.0 intellectual property (IP) and promising an “order of magnitude” improvement in compute density through TENNs support. This transformative leap underscores the inexorable shift toward Multimodal Edge AI, a trend intensifying the demands for intelligent computing at the Edge. Researchers applaud this development, highlighting that BrainChip’s second-generation Akida aligns precisely with the critical requirements of performance, efficiency, accuracy, and reliability, which are essential for expediting this transition.
At the heart of the Akida 2.0 platform lie the TENNs, offering a remarkable “order of magnitude” reduction in model size and computational demands. This leap in efficiency bodes well for accelerating AI adoption and holds the potential to make Edge AI solutions more accessible and straightforward to deploy.
As BrainChip opens the doors to the Akida 2.0 IP, eager innovators and tech enthusiasts are encouraged to connect with the company’s sales department for further information. While pricing details remain undisclosed at this juncture, the launch timeline for hardware development kits based on the second-generation platform is yet to be confirmed.
Conclusion:
BrainChip’s second-generation Akida platform represents a pivotal advancement in Edge AI. With TENN support and enhanced efficiency, it aligns perfectly with the demands of the market, offering a transformative solution for accelerating AI adoption at the Edge. This development is poised to make Edge AI solutions more accessible and signifies a significant step forward in the ever-evolving AI landscape.
