A Reconfigurable Depth-Wise Convolution Module for Heterogeneously Quantized DNNs

In Deep Neural Networks (DNN), the depth-wise separable convolution has often replaced the standard 2D convolution having much fewer parameters and operations. Another common technique to squeeze DNNs is heterogeneous quantization, which uses a different bitwidth for each layer. In this context we propose for the first time a novel Reconfigurable Depth-wise convolution Module (RDM), which uses multipliers that can be reconfigured to support 1, 2 or 4 operations at the same time at increasingly lower precision of the operands. We leveraged High Level Synthesis to produce five RDM variants with different channels parallelism to cover a wide range of DNNs. The comparisons with a non-configurable Standard Depth-wise convolution module (SDM) on a CMOS FDSOI 28-nm technology show a significant latency reduction for a given silicon area for the low-precision configurations.