Por favor, use este identificador para citar o enlazar este ítem:
https://doi.org/10.1145/3531011
Twittear
Título: | STIFT: A Spatio-Temporal Integrated Folding Tree for Efficient Reductions in Flexible DNN Accelerators |
Fecha de publicación: | 8-sep-2023 |
Editorial: | Association for Computing Machinery (ACM) |
Cita bibliográfica: | ACM Journal on Emerging Technologies in Computing Systems, Vol. 19, No. 4, Septiembre 2023. |
ISSN: | Print: 1550-4832 Electronic: 1550-4840 |
Palabras clave: | Deep Neural Networks DNN Accelerators Computer Architecture Networks-On-Chip |
Resumen: | Increasing deployment of Deep Neural Networks (DNNs) recently fueled interest in the development of specific accelerator architectures capable of meeting their stringent performance and energy consumption requirements. DNN accelerators can be organized around three separate NoCs, namely distribution, multiplier, and reduction networks (or DN, MN, and RN, respectively) between the global buffer(s) and the compute units (multipliers/adders). Among them, the RN, used to generate and reduce the partial sums produced during DNN processing, is a first-order driver of the area and energy efficiency of the accelerator. RNs can be orchestrated to exploit a Temporal, Spatial or Spatio-Temporal reduction dataflow. Among these, Spatio-Temporal reduction is the one that has shown superior performance. However, as we demonstrate in this work, a state-of-the-art implementation of the Spatio-Temporal reduction dataflow, based on the addition of Accumulators (Ac) to the RN (i.e., RN+Ac strategy), can result into significant area and energy expenses. To cope with this important issue, we propose STIFT (that stands for Spatio-Temporal Integrated Folding Tree) that implements the Spatio-Temporal reduction dataflow entirely on the RN hardware substrate (i.e., without the need for the extra accumulators). STIFT results into significant area and power savings regarding the more complex RN+Ac strategy, at the same time its performance advantage is preserved. |
Autor/es principal/es: | Muñoz-Martínez, Francisco Abellán, José L. Acacio, Manuel E. Krishna, Tushar |
Facultad/Departamentos/Servicios: | Facultades, Departamentos, Servicios y Escuelas::Departamentos de la UMU::Ingeniería y Tecnología de Computadores |
Versión del editor: | https://dl.acm.org/doi/10.1145/3531011 |
URI: | http://hdl.handle.net/10201/137602 |
DOI: | https://doi.org/10.1145/3531011 |
Tipo de documento: | info:eu-repo/semantics/article |
Número páginas / Extensión: | 19 |
Derechos: | info:eu-repo/semantics/openAccess Atribución 4.0 Internacional |
Descripción: | © 2023. Association for Computing Machinery (ACM) . This document is made available under the CC-BY 4.0 license http://creativecommons.org/licenses/by /4.0/ This document is the published version of a published work that appeared in final form in ACM Journal on Emerging Technologies in Computing Systems (JETC) |
Aparece en las colecciones: | Artículos: Ingeniería y Tecnología de Computadores |
Ficheros en este ítem:
Fichero | Descripción | Tamaño | Formato | |
---|---|---|---|---|
_JETC_Camera_Ready__RENIF_Project.pdf | 1,14 MB | Adobe PDF | Visualizar/Abrir |
Este ítem está sujeto a una licencia Creative Commons Licencia Creative Commons