Traditionally this type of capability would be deployed either through a powerful (and costly) applications processor or a microcontroller with additional components to accelerate key capabilities, says XMOS. The new xcore.ai crossover processor is architected to deliver real-time inferencing and decisioning at the edge, as well as signal processing, control and communications. With its fast processing and neural network capabilities, xcore.ai enables data to be processed locally and actions taken on device, within nanoseconds.
The processor is fully programmable in ‘C’, with specific features such as DSP and machine learning accessible through optimised c-libraries. It supports the FreeRTOS real-time operating system, enabling developers to use a broad range of familiar open-source library components. The TensorFlow Lite to xcore.ai converter allows easy prototyping and deployment of neural network models. Built with 16 real-time logical cores with support for scalar/float/vector instructions, the highly configurable processor features up to 128 pins of flexible IO (programmable in software) for access to a wide variety of interfaces and peripherals. Integrated hardware USB 2.0 PHY and MIPI interfaces allow the collection and processing of data from a wide range of sensors. The device employs deep neural networks using binary values for activations and weights instead of full precision values, dramatically reducing execution time. It is claimed to deliver 2.6x to 4x more efficiency than its 8-bit counterpart, a 32x improvement in AI performance and a 5x digital signal processing performance improvement compared to the nearest comparable ARM Cortex product.
More information at xcore.ai