Amol Borkar, Director of Product and Marketing for Vision and AI DSPs at Cadence Tensilica, demonstrates the company’s latest edge AI and vision technologies and products at the 2022 Embedded Vision Summit. Specifically, Borkar demonstrates Tensilica’s latest processor from the Vision DSP family, the Vision P1 DSP, running on Cadence’s own FPGA platform.
The Vision P1 is a highly efficient, configurable and extensible vision processor targeted for ultra-low power always-on and other applications. In this demonstration, the Vision P1 DSP is executing visual wake word and keyword spotting AI networks live and in real-time. These networks, available in the MLPerf Tiny v0.7 benchmark suite, are often used to simulate human presence detection applications. Leveraging its scalable architecture and comprehensive software libraries, the Vision P1 DSP is fully capable of handling even heavier computer vision, image processing and AI network workloads used in high performance edge applications.