STONNE: Enabling Cycle-Level Microarchitectural Simulation for DNN Inference Accelerators

The design of specialized architectures for accelerating the inference procedure of Deep Neural Networks (DNNs) is a booming area of research nowadays. While first-generation rigid accelerator proposals used simple fixed dataflows tailored for dense DNNs, more recent architectures have argued for flexibility to efficiently support a wide variety of layer types, dimensions, and sparsity. As the complexity of these accelerators grows, the analytical models currently being used for design-space exploration are unable to capture execution-time subtleties, leading to inexact results in many cases as we demonstrate. This opens up a need for cycle-level simulation tools to allow for fast and accurate design-space exploration of DNN accelerators, and rapid quantification of the efficacy of architectural enhancements during the early stages of a design. To this end, we present STONNE (Simulation TOol of Neural Network/Engines), a cycle-level microarchitectural simulation framework that can plug into any high-level DNN framework as an accelerator device and perform full-model evaluation (i.e. we are able to simulate real, complete, unmodified DNN models) of state-of-the-art rigid and flexible DNN accelerators, both with and without sparsity support. As a proof of concept, we use STONNE in three use cases: i) a direct comparison of three dominant inference accelerators using real DNN models; ii) back-end extensions and iii) front-end extensions of the simulator to showcase the capability of STONNE to rapidly and precisely evaluate data-dependent optimizations.