Range-Lookup Approximate Computing Acceleration for Any Activation Functions in Low-Power Neural Network
Wen-Chang Yang, Shuyun Lin, Tsung‐Chu Huang · 2020
Consumer electronics have become versatile for processing a lot of signals in any distribution. This results in that high-speed activating of neural network should fit for any distribution of error functions. In this paper, we propose a set of transistor-level magnitude-comparators. Then we apply them to develop range-addressable memory to design a lightweight-slope lookup table. We then develop an efficient algorithm for constructing a lightweight-slope piecewise line. The proposed techniques are suitable to design any sigmoidal activation functions in neural network. From experiments and comparisons, the proposed LUT can be more efficient effective than previous looking tables. Especially applied in the proposed range-addressable memory, the power-delay product can be reduced by more than 30 folds.