1. Deep gradient compression: Reducing the communication bandwidth for distributed training;lin;International Conference on Learning Representations,2018
2. Sparse Communication for Distributed Gradient Descent
3. A High-Throughput Low-Power Soft Bit-Flipping LDPC Decoder in 28 nm FD-SOI
4. Terngrad: Ternary gradients to reduce communication in distributed deep learning;wen;Proc Neural Inf Process Syst,2017
5. QSGD: Communication-efficient SGD via randomized quantization and encoding;alistarh;Proc Neural Inf Process Syst,2018