Microchip Technology Inc. has introduced the VectorBlox Accelerator Software Development Kit (SDK) to meet the growing need for power-efficient inferencing in edge application by making it easier for software developers to implement their algorithms in PolarFire field-programmable gate arrays (FPGAs). By using the SDK developers can use the PolarFire FPGAs for creating low-power, flexible overlay-based neural network applications without learning an FPGA tool flow.
The VectorBlox Accelerator SDK is designed to allow developers to code in C/C++ and program power-efficient neural networks and it can execute models in TensorFlow and the open neural network exchange (ONNX) format. The VectorBlox Accelerator SDK can be operated on Linux and Windows operating system, and it has a bit accurate simulator that can be used to validate the accuracy of the hardware in the software environment itself. With the help of the neural network IP, the kit also supports the ability to load different network models at run time.
The PolarFire FPGAs delivers up to 50 percent lower total power than competing devices, while also offering 25 percent higher-capacity math blocks that can deliver up to 1.5 tera operations per second (TOPS). By using the FPGAs, the developers can leverage customization and differentiation through the devices’ inherent upgradability and ability to integrate functions on a single chip. The PolarFire is available in a range of sizes for matching the performance, power, and package size tradeoffs for applications, to enable the customers to implement their solutions in package sizes as small as 11x11 mm.
To know more about the VectorBlox Accelerator SDK, visit the official website of Microchip Technology Inc.