Accelerating AI algorithms
summary
Intel announces low-power vision accelerators and a software path to move models from x86 to VPUs and FPGAs, plus a USB Intel Neural Compute Stick for local prototyping. The OpenVINO toolkit optimizes models and targets CPU, VPU or FPGA and reportedly yields up to 100x faster performance on existing CPUs. Claims include edge inference growing faster than data-center AI (25–35% annual value growth), real-time helmet detection, retail shelf pick counts, and low-power medical image inference.
AI-written summary · about
chapters (4)
related talks
24mTools & Techniques for a cost optimized Analytics & AI pipeline
50mAI acceleration at the edge
05mAccelerate AI development from cloud to client
20mComputer Vision at the Edge with OpenVINO
50mLakshmi Narasimhan of Intel at Machine Learning Developers Summit 2019
09mTackling edge computing with Intel's AI Builders program
30mAI: unleashing the next wave
22mLeverage AI Kits for Faster AI Solutions Tailored to Your Needs