Steve10
Regular
Akida™ Development Environment (ADE).I’ve been reading some more again from the end of 2019 company progress update and I can’t work out what ADE stands for. Anyone know?
Talk about intellectual property licensing a bit more. There were a lot of questions about it, and I think in part that's because we've voiced a strong opinion, coming in advance of actual device sales. There's no manufacturing process involved. There's no inventory. There's no loan package qualification by the customer. We released that in 2019. We have received strong response from prospective customers. The ADE, in the hands of one major South Korean company, is being exercised almost as much as we exercise it. They really have dug in, validated some of the benchmark results that we've provided, and now they're moving onto some of their own proprietary networks to do validation.
BrainChip’s Akida Development Environment Now Freely Available for Use
Develop and Deploy on Akida Deeply Learned Neural Networks in a standard TensorFlow/Keras EnvironmentSAN FRANCISCO–(BUSINESS WIRE)–
BrainChip Holdings Ltd. (ASX: BRN), a leading provider of ultra-low power, high-performance edge AI technology, today announced that access to its Akida™ Development Environment (ADE) no longer requires pre-approval, now allowing designers to freely develop systems for edge and enterprise products on the company’s Akida Neural Processing technology.
ADE is a complete, industry-standard machine learning framework for creating, training and testing deeply learned neural networks. The platform leverages TensorFlow and Keras for neural network development, optimization and training. Once the network model is fully trained, the ADE includes a simple-to-use compiler to map the network to the Akida fabric and run hardware accurate simulations on the Akida Execution Engine. The framework uses the Python scripting language and its associated tools and libraries, including Jupyter notebooks, NumPy and Matplotlib. With just a few lines, developers can easily run the Akida simulator on industry-standard datasets and benchmarks in the Akida model zoo such as Imagenet1000, Google Speech Commands, MobileNet among others. Users can easily create, modify, train and test their own models within a simple use development environment.
ADE comprises three main Python packages:
- the Akida Execution Engine including the Akida Simulator is an interface to the BrainChip Akida neural processing hardware. To allow the development, optimization and testing of Akida models, it includes a software backend that simulates the Akida NSoC. The output of the Akida Execution Engine generates all necessary files to run the Akida neural processor hardware as well.
- the CNN development tool utilizes TensorFlow/Keras to develop, optimize and train deeply learned neural networks such as CNNs
- the Akida model zoo contains pre-created neural network models built with the Akida sequential API and the CNN development tool using quantized Keras models.
Akida is available as a licensable IP technology that can be integrated into ASIC devices and will be available as an integrated SoC, both suitable for applications such as surveillance, advanced driver assistance systems (ADAS), autonomous vehicles (AV), vision guided robotics, drones, augmented and virtual reality (AR/VR), acoustic analysis, and Industrial Internet-of-Things (IoT). Akida is a complete neural processing engine for edge applications, which eliminates CPU and memory overhead while delivering unprecedented efficiency, faster results, at minimum cost. Functions like training, learning, and inferencing are orders of magnitude more efficient with Akida.
Access to ADE is currently available online at https://doc.brainchipinc.com/. Among the resources are installation information, user guide, API reference, Akida examples, support and license documentation. ADE requires TensorFlow 2.0.0. Any existing virtual environment previously used would need to be updated as per the installation step.