STONNE: enabling cycle-level microarchitectural simulation for DNN Inference accelerators

Muñoz Martínez, Francisco; Abellán Miguel, José Luis; Acacio Sánchez, Manuel Eugenio; Krishna, Tushar

Publication:
STONNE: enabling cycle-level microarchitectural simulation for DNN Inference accelerators

dc.contributor.author	Muñoz Martínez, Francisco
dc.contributor.author	Abellán Miguel, José Luis
dc.contributor.author	Acacio Sánchez, Manuel Eugenio
dc.contributor.author	Krishna, Tushar
dc.contributor.department	Ingeniería y Tecnología de Computadores
dc.contributor.other	Facultad de Informática
dc.date.accessioned	2026-02-26T08:32:05Z
dc.date.available	2026-02-26T08:32:05Z
dc.date.copyright	© 2022 IEEE
dc.date.issued	2022-01-13
dc.description.abstract	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.
dc.format	application/pdf
dc.format.extent	4
dc.identifier.citation	F. Muñoz-Martínez, J. L. Abellán, M. E. Acacio and T. Krishna, "STONNE: Enabling Cycle-Level Microarchitectural Simulation for DNN Inference Accelerators," 2021 IEEE International Symposium on Workload Characterization (IISWC), Storrs, CT, USA, 2021, pp. 201-213, doi: 10.1109/IISWC53511.2021.00028
dc.identifier.doi	https://doi.org/10.1109/IISWC53511.2021.00028
dc.identifier.eisbn	978-1-6654-4173-5
dc.identifier.isbn	978-1-6654-4174-2
dc.identifier.uri	http://hdl.handle.net/10201/214001
dc.language	eng
dc.publisher	IEEE
dc.relation	Work supported by RTI2018-098156-B-C53 (MCIU/AEI/FEDER,UE), NSFOAC 1909900 and US Department of Energy ARIAA co-design center. F. Munoz-Matíınez supported by grant 20749/FPI/18 from Fundacion Séneca.
dc.relation.ispartof	2021 IEEE International Symposium on Workload Characterization (IISWC), 7-9 Nov. 2021
dc.relation.publisherversion	https://ieeexplore.ieee.org/document/9668279
dc.rights	Attribution-NonCommercial-NoDerivatives 4.0 International	*
dc.rights.accessRights	info:eu-repo/semantics/openAccess
dc.rights.uri	http://creativecommons.org/licenses/by-nc-nd/4.0/	*
dc.subject	Simulation tool
dc.subject	Specialized architectures for DNNs
dc.subject	DNN Accelerators
dc.subject.ods	No relacionado con ningún objetivo de desarrollo sostenible
dc.title	STONNE: enabling cycle-level microarchitectural simulation for DNN Inference accelerators
dc.type	info:eu-repo/semantics/conferenceObject
dc.type.version	info:eu-repo/semantics/acceptedVersion
dspace.entity.type	Publication	es
relation.isAuthorOfPublication	dbf4e7f7-682b-4a97-a941-16178079ac7a
relation.isAuthorOfPublication	0408fa9f-b61e-48cb-a08f-ae8999f885ee
relation.isAuthorOfPublication	5aeaee45-9977-413f-94ee-2a27e25672f8
relation.isAuthorOfPublication.latestForDiscovery	dbf4e7f7-682b-4a97-a941-16178079ac7a