Fast Linear Interpolation for Piecewise-Linear Functions, GAMs, and Deep Lattice Networks

Nathan Zhang, Kevin Canini, Sean Silva, and Maya R. Gupta

Keywords: hardware, compiler, MLIR, runtime, CPU, interpolation

TL;DR: Fast implementations of linear interpolation operators are given for both piecewise linear functions and multi-dimensional look-up tables, producing 3-11x faster runtimes for single evaluations.

Abstract: We present fast implementations of linear interpolation operators for both piecewise linear functions and multi-dimensional look-up tables. We use a compiler-based solution (using MLIR) for accelerating this family of workloads. On real-world multi-layer lattice models and a standard CPU, we show these strategies deliver $5-10\times$ faster runtimes compared to a C++ interpreter implementation that uses prior techniques, producing runtimes that are 1000s of times faster than TensorFlow 2.0 for single evaluations.