Lecture 10: Parallel Architectures for Inference Challenges of inference, low-bit representations, pruning, GPU vs FPGA and ASIC, TPU architecture. Slides Video Previous Next