Communication Dans Un Congrès Année : 2021

Iterative Method for Performance Prediction Improvement of Integrated Circuits

Résumé

Currently, the design flow of digital integrated circuits relies on the use of pre-characterized libraries. These libraries, pre-calculated outside of any context, are used at each step of the design flow to predict the performance of the circuit. They are necessary for the calculation of a wide variety of characteristics, including the IR Drop. However, the behavior of the logic gates described in these libraries is not representative of the real power supply of the standard-cells and this difference is even more important as the number of logic gates in the circuits increases. In this paper, we will show that with the current design flow, the use of these libraries leads to an underestimation of the circuit performances. To remedy this, we propose an iterative approach based on a new use of voltageadaptive libraries that allows to compute the IR Drop more precisely and thus to better predict the performances of the circuits.

Fichier principal
Vignette du fichier
paper21-final.pdf (839.18 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

lirmm-03710383 , version 1 (30-06-2022)

Licence

Identifiants

Citer

Gwenaël Chaillou, Philippe Maurine, Jean-Marc J.-M. Galliere, Nadine Azemard. Iterative Method for Performance Prediction Improvement of Integrated Circuits. DCIS 2021 - 36th Conference on Design of Circuits and Integrated Systems, Nov 2021, Vila do Conde, Portugal. pp.1-5, ⟨10.1109/DCIS53048.2021.9666182⟩. ⟨lirmm-03710383⟩
242 Consultations
417 Téléchargements

Altmetric

Partager

  • More