From Specification to Silicon: Towards Analog/Mixed-Signal Design Automation using Surrogate NN Models with Transfer Learning

Juzheng Liu; Shiyu Su; Meghna Madhusudan; Mohsen Hassanpourghadi; Samuel Saunders; Qiaochu Zhang; Rezwan Rasul; Yaguang Li; Jiang Hu; Arvind Kumar Sharma; Sachin S. Sapatnekar; Ramesh Harjani; Anthony Levi; Sandeep Gupta; Mike Shuo Wei Chen

doi:10.1109/ICCAD51958.2021.9643445

From Specification to Silicon: Towards Analog/Mixed-Signal Design Automation using Surrogate NN Models with Transfer Learning

Juzheng Liu, Shiyu Su, Meghna Madhusudan, Mohsen Hassanpourghadi, Samuel Saunders, Qiaochu Zhang, Rezwan Rasul, Yaguang Li, Jiang Hu, Arvind Kumar Sharma, Sachin S. Sapatnekar, Ramesh Harjani, Anthony Levi, Sandeep Gupta, Mike Shuo Wei Chen

Electrical and Computer Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

9 Scopus citations

Abstract

We propose a complete analog mixed-signal circuit design flow from specification to silicon with minimum human-in-the-loop interaction, and verify the flow in a 12nm FinFET CMOS process. The flow consists of three key elements: neural network (NN) modeling of the parameterized circuit component, a search algorithm based on NN models to determine its sizing, and layout automation. To reduce the required training data for NN model creation, we utilize transfer learning to improve the NN accuracy from a relatively small amount of post-layout/silicon data. To prove the concept, we use a voltage-controlled oscillator (VCO) as a test vehicle and demonstrate that our design methodology can accurately model the circuit and generate designs with a wide range of specifications. We show that circuit sizing based on the transfer learned NN model from silicon measurement data yields the most accurate results.

Original language	English (US)
Title of host publication	2021 40th IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2021 - Proceedings
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781665445078
DOIs	https://doi.org/10.1109/ICCAD51958.2021.9643445
State	Published - 2021
Event	40th IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2021 - Munich, Germany Duration: Nov 1 2021 → Nov 4 2021

Publication series

Name	IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD
Volume	2021-November
ISSN (Print)	1092-3152

Conference

Conference	40th IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2021
Country/Territory	Germany
City	Munich
Period	11/1/21 → 11/4/21

Bibliographical note

Funding Information:
The authors wish to acknowledge support from the DARPA POSH program (FA8650-18-2-7853) and program manager Serge Leef. We also thank Global Foundries for access to GF12LP technology. Qiaochu Zhang also acknowledges the funding support from the University of Southern California Provost’s Fellowship.

Funding Information:
The authors wish to acknowledge support from the DARPA POSH program (FA8650-18-2-7853) and program manager Serge Leef. We also thank Global Foundries for access to GF12LP technology. Qiaochu Zhang also acknowledges the funding support from the University of Southern California Provost?s Fellowship.

Publisher Copyright:
© 2021 IEEE

Keywords

AMS circuit design automation
Circuit modeling
Layout automation
Silicon verified CAD

Publisher link

10.1109/ICCAD51958.2021.9643445

Find It

Find It at U of M Libraries

Cite this

Liu, J., Su, S., Madhusudan, M., Hassanpourghadi, M., Saunders, S., Zhang, Q., Rasul, R., Li, Y., Hu, J., Sharma, A. K., Sapatnekar, S. S., Harjani, R., Levi, A., Gupta, S., & Chen, M. S. W. (2021). From Specification to Silicon: Towards Analog/Mixed-Signal Design Automation using Surrogate NN Models with Transfer Learning. In 2021 40th IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2021 - Proceedings (IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD; Vol. 2021-November). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICCAD51958.2021.9643445

From Specification to Silicon: Towards Analog/Mixed-Signal Design Automation using Surrogate NN Models with Transfer Learning. / Liu, Juzheng; Su, Shiyu; Madhusudan, Meghna et al.
2021 40th IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2021 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2021. (IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD; Vol. 2021-November).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Liu, J, Su, S, Madhusudan, M, Hassanpourghadi, M, Saunders, S, Zhang, Q, Rasul, R, Li, Y, Hu, J, Sharma, AK, Sapatnekar, SS , Harjani, R, Levi, A, Gupta, S & Chen, MSW 2021, From Specification to Silicon: Towards Analog/Mixed-Signal Design Automation using Surrogate NN Models with Transfer Learning. in 2021 40th IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2021 - Proceedings. IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD, vol. 2021-November, Institute of Electrical and Electronics Engineers Inc., 40th IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2021, Munich, Germany, 11/1/21. https://doi.org/10.1109/ICCAD51958.2021.9643445

Liu J, Su S, Madhusudan M, Hassanpourghadi M, Saunders S, Zhang Q et al. From Specification to Silicon: Towards Analog/Mixed-Signal Design Automation using Surrogate NN Models with Transfer Learning. In 2021 40th IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2021 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2021. (IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD). doi: 10.1109/ICCAD51958.2021.9643445

Liu, Juzheng ; Su, Shiyu ; Madhusudan, Meghna et al. / From Specification to Silicon : Towards Analog/Mixed-Signal Design Automation using Surrogate NN Models with Transfer Learning. 2021 40th IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2021 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2021. (IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD).

@inproceedings{ec060fb5a90b4f28a0d8a1f8cc79dd52,

title = "From Specification to Silicon: Towards Analog/Mixed-Signal Design Automation using Surrogate NN Models with Transfer Learning",

abstract = "We propose a complete analog mixed-signal circuit design flow from specification to silicon with minimum human-in-the-loop interaction, and verify the flow in a 12nm FinFET CMOS process. The flow consists of three key elements: neural network (NN) modeling of the parameterized circuit component, a search algorithm based on NN models to determine its sizing, and layout automation. To reduce the required training data for NN model creation, we utilize transfer learning to improve the NN accuracy from a relatively small amount of post-layout/silicon data. To prove the concept, we use a voltage-controlled oscillator (VCO) as a test vehicle and demonstrate that our design methodology can accurately model the circuit and generate designs with a wide range of specifications. We show that circuit sizing based on the transfer learned NN model from silicon measurement data yields the most accurate results.",

keywords = "AMS circuit design automation, Circuit modeling, Layout automation, Silicon verified CAD",

author = "Juzheng Liu and Shiyu Su and Meghna Madhusudan and Mohsen Hassanpourghadi and Samuel Saunders and Qiaochu Zhang and Rezwan Rasul and Yaguang Li and Jiang Hu and Sharma, {Arvind Kumar} and Sapatnekar, {Sachin S.} and Ramesh Harjani and Anthony Levi and Sandeep Gupta and Chen, {Mike Shuo Wei}",

note = "Funding Information: The authors wish to acknowledge support from the DARPA POSH program (FA8650-18-2-7853) and program manager Serge Leef. We also thank Global Foundries for access to GF12LP technology. Qiaochu Zhang also acknowledges the funding support from the University of Southern California Provost{\textquoteright}s Fellowship. Funding Information: The authors wish to acknowledge support from the DARPA POSH program (FA8650-18-2-7853) and program manager Serge Leef. We also thank Global Foundries for access to GF12LP technology. Qiaochu Zhang also acknowledges the funding support from the University of Southern California Provost?s Fellowship. Publisher Copyright: {\textcopyright} 2021 IEEE; 40th IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2021 ; Conference date: 01-11-2021 Through 04-11-2021",

year = "2021",

doi = "10.1109/ICCAD51958.2021.9643445",

language = "English (US)",

series = "IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2021 40th IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2021 - Proceedings",

}

TY - GEN

T1 - From Specification to Silicon

T2 - 40th IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2021

AU - Liu, Juzheng

AU - Su, Shiyu

AU - Madhusudan, Meghna

AU - Hassanpourghadi, Mohsen

AU - Saunders, Samuel

AU - Zhang, Qiaochu

AU - Rasul, Rezwan

AU - Li, Yaguang

AU - Hu, Jiang

AU - Sharma, Arvind Kumar

AU - Sapatnekar, Sachin S.

AU - Harjani, Ramesh

AU - Levi, Anthony

AU - Gupta, Sandeep

AU - Chen, Mike Shuo Wei

N1 - Funding Information: The authors wish to acknowledge support from the DARPA POSH program (FA8650-18-2-7853) and program manager Serge Leef. We also thank Global Foundries for access to GF12LP technology. Qiaochu Zhang also acknowledges the funding support from the University of Southern California Provost’s Fellowship. Funding Information: The authors wish to acknowledge support from the DARPA POSH program (FA8650-18-2-7853) and program manager Serge Leef. We also thank Global Foundries for access to GF12LP technology. Qiaochu Zhang also acknowledges the funding support from the University of Southern California Provost?s Fellowship. Publisher Copyright: © 2021 IEEE

PY - 2021

Y1 - 2021

N2 - We propose a complete analog mixed-signal circuit design flow from specification to silicon with minimum human-in-the-loop interaction, and verify the flow in a 12nm FinFET CMOS process. The flow consists of three key elements: neural network (NN) modeling of the parameterized circuit component, a search algorithm based on NN models to determine its sizing, and layout automation. To reduce the required training data for NN model creation, we utilize transfer learning to improve the NN accuracy from a relatively small amount of post-layout/silicon data. To prove the concept, we use a voltage-controlled oscillator (VCO) as a test vehicle and demonstrate that our design methodology can accurately model the circuit and generate designs with a wide range of specifications. We show that circuit sizing based on the transfer learned NN model from silicon measurement data yields the most accurate results.

AB - We propose a complete analog mixed-signal circuit design flow from specification to silicon with minimum human-in-the-loop interaction, and verify the flow in a 12nm FinFET CMOS process. The flow consists of three key elements: neural network (NN) modeling of the parameterized circuit component, a search algorithm based on NN models to determine its sizing, and layout automation. To reduce the required training data for NN model creation, we utilize transfer learning to improve the NN accuracy from a relatively small amount of post-layout/silicon data. To prove the concept, we use a voltage-controlled oscillator (VCO) as a test vehicle and demonstrate that our design methodology can accurately model the circuit and generate designs with a wide range of specifications. We show that circuit sizing based on the transfer learned NN model from silicon measurement data yields the most accurate results.

KW - AMS circuit design automation

KW - Circuit modeling

KW - Layout automation

KW - Silicon verified CAD

UR - http://www.scopus.com/inward/record.url?scp=85121629564&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85121629564&partnerID=8YFLogxK

U2 - 10.1109/ICCAD51958.2021.9643445

DO - 10.1109/ICCAD51958.2021.9643445

M3 - Conference contribution

AN - SCOPUS:85121629564

T3 - IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD

BT - 2021 40th IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2021 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 1 November 2021 through 4 November 2021

ER -

From Specification to Silicon: Towards Analog/Mixed-Signal Design Automation using Surrogate NN Models with Transfer Learning

Abstract

Publication series

Conference

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this