Compact TinyBeast FPGA board to accelerate edge AI
2024-05-05 10:13
Sundance DSP has announced the release of a compact board called TinyBeast FPGA, based on Microchip's PolarFire field programmable gate array (FPGA).
This device is designed to accelerate embedded and industrial computers and provides full support for artificial intelligence (AI) acceleration using Microchip's VectorBlox technology.
The TinyBeast FPGA is available in two variants: a mini-PCI Express (mPCIe) board and a surface mount module. Both variants are based on Microchip's PolarFire MPF300T-1FCVG484E FPGA, which offers high performance including 300,000 logic elements, 924 math blocks and 20.6 MB of total RAM. The device also features 16 SERDES lanes, two PCI Express endpoints and up to 512 user-accessible input/output (IO). However, the main feature of the TinyBeast FPGA is its ability to accelerate edge AI and machine learning workloads using Microchip's VectorBlox technology.
VectorBlox is a hardware gas pedal designed specifically for processing AI inference. This enables applications such as predictive maintenance, anomaly detection and image recognition right at industrial sites.