BrainChip’s Cloud Platform Lets Developers Use Its Neuromorphic Tech

BrainChip has launched Developer Akida Cloud, a cloud-based access point to the company's neuromorphic technology, designed to reduce the time and effort required to use its Akida Cloud platform.

To accelerate product development and deliver reliable applications, developers can leverage this platform to program and run their network models. Meanwhile, Akida's second-generation technology, in its latest version (Akida 2), will be included in the Developer Cloud presentation.

Developer Akida Cloud is a new cloud-based access point to multiple generations and configurations of BrainChip’s Akida neuromorphic technology.

Helping Developers Access Neuromorphic Technology

Developers can benefit from Akida Cloud, accessing numerous configurations and generations of BrainChip's Akida neuromorphic technology from the cloud. According to the company, Akida products are quickly and easily delivered to developers on a roadmap, giving them the possibility of having the latest Akida technology to deliver reliable developments.

Akida Cloud helps developers achieve product design in less time by making it easier to program and run their network models. These tools enable the deployment of advanced AI applications across multiple industries, with a simplified development process and reduced model size.

The premise of this cloud platform is rapid application development. Immediate results are achieved through direct access to Akida's latest technology, making it easy to program and run neural networks and develop an idea from concept to product. 

Akida Cloud Features

Akida Cloud enables real-time data submission so developers can perform inferences. This platform accelerates innovation through its practical testing capabilities. Users gain access to Akida's latest capabilities, allowing them to advance the development of their neuromorphic models by remotely testing innovations.

BrainChip deploys new versions and configurations of Akida via the cloud, so developers can stay up-to-date with the latest resources. Akida2 now holds up 8-bit quantization, allowing developers to create more sophisticated models with greater accuracy.

If you want to process temporal video, audio, and sensor data, Akida 2-compatible models improve performance by incorporating event-based temporal neural networks (TENNs). These networks are suitable for real-time streaming applications based on events occurring in state-space models (SSM).

The data is sent by the developer to Akida Cloud in real time, allowing them to perform inferences on their computer. The cloud platform sends the results back, facilitating iteration and rapid training, resulting in a more accurate model in real-world conditions. This allows applications to be developed through a systematic process of data acquisition, processing, and visualization.

Akida Technology for Hardware Development

Akida 2 is the latest version of Akida's second-generation technology to be included in Developer Cloud, designed with AI efficiency and low power consumption as a priority. Compared to Akida 1, it offers a four-times increase in performance and efficiency.

This platform allows developers to utilize the most innovative features for rapid prototyping without physical hardware. This way, engineers can perform hardware deployment alongside cloud development. Developers also have access to an FPGA platform through Akida for a local solution.

To understand Akida's capabilities, you can find several applications in Jupyter Labs, a web-based interactive development environment (IDE), which can be run on Akida 2.

Neuromorphic chip: Akida

Boosting Profits for R&D Teams

Industries that can develop applications utilizing low-power edge AI include aerospace, robotics, consumer devices, autonomous vehicles, industrial IoT, and wearables. With the goal of boosting the development of intelligent computing, BrainChip proposes bringing AI closer to the sensor and real-time.

R&D teams benefit from the ability this platform offers to remotely test their neuromorphic models with the goal of further refining them. This platform makes it possible to customize its configuration to suit the user's requirements for integration into an IC.

Perspectives for the Future

To advance in the field of Artificial Intelligence, it is essential to have efficient tools that allow us to keep up with its rapid growth. This platform enables the development of innovative intelligent edge devices with simple-to-deploy models and compact sizes.

The benefits of making new versions of Akida available through the cloud are reflected in the different node configurations available, as well as increased memory and system design improvements.

The goal for the future is for Akida to be able to meet the demanding needs of new AI applications, being able, as in the case of Akida Cloud, to provide solutions under high workload conditions. 

All Images used courtesy of BrainChip.