Presentation | 2020-01-17 [Poster Presentation] Investigation of Placement Order Optimization for Adiabatic Quantum- Flux-Parametron Integrated Circuits via Machine Learning Takehisa Yamada, Christopher L. Ayala, Ro Saito, Tomoyuki Tanaka, Nobuyuki Yoshikawa, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Adiabatic quantum-flux-parametron (AQFP) logic is one kind of superconducting logic family featuring low energy and high computational speed compared to CMOS. Because of its unique structure, we cannot immediately apply CMOS EDA tools to AQFP circuits. In previous studies, we developed an AQFP cell placement optimization tool using the genetic algorithm (GA). However, GA-based optimization time is enormous when it is applied to very large integrated circuits (IC). In this study, we have been focusing on optimization by machine learning. Once such an optimization model has been created, an optimized placement result can be obtained very quickly by utilizing the model. We first generated pseudo-circuit data, and we trained an order optimization model using the pseudo-circuit data. To apply machine learning to the placement optimization problem, a series data must be created from the AQFP graph data. In this paper, we propose the following two methods. In the first method, we create feature vectors for each AQFP cell from the graph data to make the series data. In the second method, we optimize the placement order by applying the vectors in machine learning. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | AQFPplacementoptimizationmachine learning |
Paper # | SCE2019-55 |
Date of Issue | 2020-01-09 (SCE) |
Conference Information | |
Committee | SCE |
---|---|
Conference Date | 2020/1/16(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | Satoshi Kohjiro(AIST) |
Vice Chair | |
Secretary | (Yokohama National Univ.) |
Assistant | Hiroyuki Akaike(Daido Univ.) |
Paper Information | |
Registration To | Technical Committee on Superconductive Electronics |
---|---|
Language | ENG |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | [Poster Presentation] Investigation of Placement Order Optimization for Adiabatic Quantum- Flux-Parametron Integrated Circuits via Machine Learning |
Sub Title (in English) | |
Keyword(1) | AQFPplacementoptimizationmachine learning |
1st Author's Name | Takehisa Yamada |
1st Author's Affiliation | Yokohama National University(Yokohama Natl Univ) |
2nd Author's Name | Christopher L. Ayala |
2nd Author's Affiliation | Yokohama National University(Yokohama Natl Univ) |
3rd Author's Name | Ro Saito |
3rd Author's Affiliation | Yokohama National University(Yokohama Natl Univ) |
4th Author's Name | Tomoyuki Tanaka |
4th Author's Affiliation | Yokohama National University(Yokohama Natl Univ) |
5th Author's Name | Nobuyuki Yoshikawa |
5th Author's Affiliation | Yokohama National University(Yokohama Natl Univ) |
Date | 2020-01-17 |
Paper # | SCE2019-55 |
Volume (vol) | vol.119 |
Number (no) | SCE-369 |
Page | pp.pp.103-105(SCE), |
#Pages | 3 |
Date of Issue | 2020-01-09 (SCE) |