3 ms·Throughput Optimizations for FPGA-Based Deep Neural Network Inference1 points by godelmachine 8y ago