Feeding Hungry Models Less: Deep Transfer Learning for Embedded Memory PPA Models : al Session

Felix Last, Ulf Schlichtmann

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

3 Scopus citations

Abstract

Supervised machine learning requires large amounts of labeled data for training. In power, performance and area (PPA) estimation of embedded memories, every new memory compiler version is considered independently of previous versions. Since the data of different memory compilers originate from similar domains, transfer learning may reduce the amount of supervised data required by pre-training PPA estimation neural networks on related domains. We show that provisioning times of PPA models for new compiler versions can be reduced significantly by exploiting similarities across versions and technology nodes. Through transfer learning, we shorten the time to provision PPA models for new compiler versions by 50% to 90%, which speeds up time-critical periods of the design cycle. This is achieved by requiring less than 6,500 ground truth samples for the target compiler to achieve average estimation errors of 0.35% instead of 13,000 samples. Using only 1,300 samples is sufficient to achieve an almost worst-case (98th percentile) error of approximately 3% and allows us to shorten model provisioning times from over 40 days to less than one week.

Original languageEnglish
Title of host publication2021 ACM/IEEE 3rd Workshop on Machine Learning for CAD, MLCAD 2021
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781665431668
DOIs
StatePublished - 30 Aug 2021
Event3rd ACM/IEEE Workshop on Machine Learning for CAD, MLCAD 2021 - Raleigh, United States
Duration: 30 Aug 20213 Sep 2021

Publication series

Name2021 ACM/IEEE 3rd Workshop on Machine Learning for CAD, MLCAD 2021

Conference

Conference3rd ACM/IEEE Workshop on Machine Learning for CAD, MLCAD 2021
Country/TerritoryUnited States
CityRaleigh
Period30/08/213/09/21

Keywords

  • artificial neural networks
  • deep learning
  • electronic design automation
  • few-shot learning
  • machine learning
  • memory compilers
  • regression
  • transfer learning

Fingerprint

Dive into the research topics of 'Feeding Hungry Models Less: Deep Transfer Learning for Embedded Memory PPA Models : al Session'. Together they form a unique fingerprint.

Cite this