Archives: AI in Wave division

The recent study on artificial neural network signal equalization authored by Ahmed Galib Reza (KAIST EE) and June-Koo Kevin Rhee has been published in IEEE Photonics Technology Letters.  ( https://ieeexplore.ieee.org/document/8401897 )

Article Content:

Title: Nonlinear Equalizer Based on Neural Networks for PAM-4 Signal Transmission Using DML

Authors: Ahmed Galib Reza, and June-Koo Kevin Rhee

Nonlinear distortion from a directly modulated laser (DML) is one of the major limiting factors to enhance the transmission capacity beyond 10 Gb/s for an intensity modulation direct-detection optical access network. In this letter, we propose and demonstrate a low-complexity nonlinear equalizer (NLE) based on a machine-learning algorithm called artificial neural network (ANN). Experimental results for a DML-based 20-Gb/s signal transmission over an 18-km SMF-28e fiber at 1310-nm employing pulse amplitude modulation (PAM)-4 confirm that the proposed ANN-NLE equalizer can increase the channel capacity and significantly reduce the impact of nonlinear penalties.

 

 

The KAIST ITRC of Quantum Computing for AI (QCAI), sponsored by the Ministry of Science and Information and led by Professor June-Koo Rhee, has been established in June 2018, to embrace strategic efforts to develop quantum HW and SW for AI technologies for the next 4 years.

Link: https://www.yna.co.kr/view/AKR20181002102000063

Professor June-Koo Rhee’s presentation at ‘2018 Pre Smart Cloud Show: Commercialization of Quantum Computing’ was reported in IT Chosun

Quantum computing is getting a lot of attention as it is an opportunity to develop the rapid computing performance that goes beyond Moore’s law. Currently, the quantum computing in the R&D stage can be implemented only in a laboratory environment. However, some of the key technologies for implementing them such as superconductivity have already been transferred to the engineering stage.

In this talk, Professor June-Koo Rhee said, “While optimism and pessimism coexist, such as the point that the current encryption system may be broken when the quantum computing is put into practical use, there is no disagreement in observation that the quantum computing will grow into the key technology for overcoming the limit at the stage of the completion of the fourth industrial revolution. In order to bring the quantum computing science to the stage of engineering, we must constantly invest in basic research.”

In this lecture, the vision and commercialization of quantum computing technology were introduced. For more information, please refer to the link below.

<Link to the article>
http://m.it.chosun.com/m/m_article.html?no=2851189

 

Title: “Machine Learning based Optimal Signal Integrity/Power Integrity Design for 3D ICs,” is published on IEEE Trans. VLSI Systems. (https://ieeexplore.ieee.org/abstract/document/7850943)

Author: Sung-Joo Park ; Bum-Hee Bae ; Joung-Ho Kim ; Madhavan Swaminathan

Article contents

Machine Learning based Optimal Signal Integrity/Power Integrity Design for 3D ICs

1. Deep Neural Network (DNN)-based Signal integrity/Power integrity Results Estimation Method

Fig. 1. Deep Neural Network (DNN)-based SI/PI results estimation method

High-speed channels and power distribution networks (PDNs) must be simulated to ensure signal/power integrity in the early stage of the entire design process for the reduction of the time and cost. However, the time to simulate the entire channels and PDNs by conventional EM, circuit simulators has been longer as the complexity of the design is increased. The estimation of the SI/PI results and model parameters using the deep neural network (DNN) can save time and cost than the conventional simulations. Because the DNN can automatically solve the non-linearity relationship between input and output, DNN can accurately estimate/model the electrical characteristics of design parameters such as high speed channels, power distribution networks, and through silicon vias (TSVs) to obtain outputs such as eye diagram, P/G noise and TSV models as shown Fig. 1. Fig. 2 shows the comparison between simulation and DNN which estimate the eye height and eye width of the high speed memory channel in HBM interposer. The estimation using DNN is accurate. As a result, the DNN can be used in many ways in the SI/PI field.

Fig. 2. Comparison between simulation and DNN model which estimate the eye height and eye width.

 

2. Reinforcement Learning-based Signal Integrity/Power Integrity Design Method

Fig. 3. Reinforcement learning (RL)-based optimal SI/PI design method.

As the demand on higher performance such as channel speed, bandwidth, and low power is increasing, the complexity of the 2.5-D/3D IC design is gradually increasing to ensure the signal/power integrity. Moreover, time-to-market is getting shorter in order to respond quickly to market trends and customer needs. Therefore, time-efficient and accurate optimal 2.5D/3D IC design is necessary in these days. Optimal layout design considering SI/PI can be performed by reinforcement learning (RL) as shown in Fig. 3.

By using RL algorithms, optimal layout design guideline which is optimal policy can be learned through reward (feedback) mechanisms depending on the target specifications of SI/PI. Therefore, high speed channels and power distribution networks (PDNs) can be designed through the RL-based optimal design method. Fig. 4 shows the results of the RL-based optimal decoupling capacitor design method. As shown in Fig. 4, the self PDN impedance of optimized PDN satisfied the target impedance and simultaneous switching noise (SSN) is suppressed.

Fig. 4. PDN self impedance and simultaneous switching noise (SSN) of the optimized PDN by the RL-based optimal design method.