[paper] Benchmarks for SPICE AI/ML Modeling

Colin C. McAndrew, Andries J. Scholten, Kiran K. Gullapalli, Yogesh Chauhan, and Kejun Xia

Benchmarks for SPICE Modeling and Parameter Extraction Based on AI/ML

IEEE Transactions on Electron Devices (2025)

DOI: 10.1109/TED.2025.3537952

1 Computer Engineering, Iowa State University, (USA)
2 NXP Semiconductors N.V., Eindhoven (NL)
3 NXP Semiconductors N.V., Austin (USA)
4 IIT Kanpur, (IN)
5 TSMC, Hsinchu (TW)

Abstract: Over the past decades, the number of submitted articles that use numerical approaches for SPICE models or for characterization (extraction) of parameters of existing SPICE models has grown significantly. Many of those articles rely on synthetic data, generated either from technology computer-aided design (TCAD) or from physical SPICE model simulations; most do not model/fit measured data. Furthermore, those articles do not evaluate the physical correctness, smoothness/monotonicity, or asymptotic correctness of the approach they propose. That is sufficient for initial evaluation of proposed techniques. However, it does not prove that they are “industrial strength.” This article presents benchmarks/guidelines for the proposed artificial intelligence (AI)/machine learning (ML) SPICE modeling and characterization techniques to try to help them become practical and useful.

TAB: CHECKLIST FOR MODELS

Capability	Existing State-of-the-Art	Proposed AI/ML Approach	Best Prior AI/ML Approach
obeys the laws of thermodynamics	✓	?	?
accurate DC modeling for all terminal currents, on relevant log/linear scale	✓	?	?
accurate capacitance/charge modeling	✓	?	?
models DC and capacitance interaction where relevant	✓	?	?
accurate modeling of high-frequency/non-quasi-static effects where relevant	✓	?	?
works for large-signal transient simulation, including delay effects	✓	?	?
accurate noise modeling	✓	?	?
has full geometry dependence	✓	?	?
has complete temperature dependence	✓	?	?
models all necessary LDEs	✓	?	?
behaves “well” for unreasonable geometry or temperature or bias	✓	?	?
exhibits physical monotonicity over bias, geometry, and temperature	✓	?	?
is smooth (ideally C∞-continuous)	✓	?	?
exhibits relevant physical symmetries (currents, charges, their derivatives)	✓	?	?
exhibits asymptotic correctness over geometry, temperature, and bias	✓	?	?
includes modeling of electrothermal effects (with frequency dependence)	✓	?	?
includes, or enables, modeling of global and local statistical variation	✓	?	?
includes, or enables, modeling of aging	✓	?	?
enables modeling of parasitics for different layouts	✓	?	?
is verified to converge reliably in at least one circuit simulator	✓	?	?

[paper] Cryo HiSIM Compact Model

Dondee Navarro, Chika Tanaka, Kyoto Institute, Shin Taniguchi, Kazutoshi Kobayashi, Michihiro Shintani and Takashi Sato

Physics-based Modeling to Extend MOSFET Compact Model for Cryogenic Operation

(Invited Paper) ASPDAC ’25, January 20–23, 2025, Tokyo, Japan

DOI: 10.1145/3658617.3703137

1 KIOXIA Corporation Yokohama (J)

2 Kyoto Institute of Technology Kyoto (J)

3 Kyoto University Kyoto (J)

Abstract: This paper extends the low-temperature modeling capabilities of an industry-standard compact metal-oxide-semiconductor field-effect transistor (MOSFET) model by incorporating physics-based representations of cryogenic effects in semiconductors. Specifically, the incomplete dopant ionization effect is integrated into the bulk Fermi potential calculation of the compact model and applied as a threshold voltage shift in the formulation of Poisson’s equation. Temperature-related models for bandgap energy, saturation velocity, and contact resistance at the source/drain regions are also enhanced. Using transistors fabricated with 22nm process technology, we demonstrate that this consistent modeling approach accurately reproduces current-voltage and threshold voltage-temperature characteristics across a temperature range from 300K to 4K.

FIG: Extracted Vth-T and SS-T characteristics from measurements

and extended HiSIM model simulations.

Acknowledgments: This work was also supported through the activities of d-lab VDEC, the University of Tokyo, in collaboration with NIHON SYNOPSYS G.K., Cadence Design Systems, and Siemens Electronic Design Automation Japan K.K.

[paper] FDSOI CMOS Cryogenic SPICE Models

P. Chava1, H. Alius2, J. Bühler1, A. R. Cabrera-Galicia1, C. Degenhardt1, T. Gneiting2, M. Harff1, T. Heide3, P. Javorka4, M. Lederer5, S. Lehmann4, M. Simon5, M. Su2, P. Vliex1, S. van Waasen1,6, C. Witt7, D. Zetzsche3

Evaluation of Cryogenic Models for FDSOI CMOS Transistors

16th IEEE Workshop on Low Temperature electronics, IEEE WOLTE16, Cagliari, Italy, Jun. 3-6, 2024

DOI: 10.34734/FZJ-2024-05369

1 Central Institute of Engineering (ZEA-2), Forschungszentrum Jülich GmbH, 52428, Jülich, (D)
2 AdMOS GmbH, 72636 Frickenhausen, (D)
3 Raycics GmbH, 01069 Dresden, (D)
4 GlobalFoundries, 01109, Dresden, (D)
5 Fraunhofer Institute for Photonic Microsystems IPMS, Center Nanoelectronic Technologies (CNT), 01109, Dresden, (D)
6 Faculty of Engineering, Communication Systems, University Duisburg-Essen, 47057 Duisburg, (D)
7 GlobalFoundries, Kapeldreef 75, 3001 Leuven, (B)

Abstract: Scalable quantum computers demand innovative solutions for tackling the wiring bottleneck to control an increasing number of qubits. Cryogenic electronics based on CMOS technologies are promising candidates which can operate down to deep-cryogenic temperatures and act as a communication and control interface to the quantum layer [1,2]. However, the performance of transistors used in these circuits is altered significantly when cooling from room temperature to cryogenic temperatures, which motivates accurate cryogenic modeling of transistors. We will report on cryogenic models tailored specifically for fully depleted silicon-on-insulator (FDSOI) transistors. We performed extensive DC characterization of transistors with subsequent modeling using the BSIM-IMG 102-9.6 model, which is the first version with a built-in cryogenic extension [3]. The preliminary models effectively represent the DC device behavior from 7 K up to room temperature. These models are used in industry standard EDA and simulation software, like Cadence Spectre. With the presented cryogenic models, we will show simulations at cryogenic temperatures. We will also compare the simulation results with the measured performance of a test chip in the temperature range from 7 K up to room temperature.

FIG: Measured and modeled transfer characteristics of a short-channel nMOST at T = 7 K

with measurement setup inside the cryogenic chamber  

Acknowledgements: This work was funded by the German Federal Ministry of Education and Research (BMBF), funding program “Quantum technologies - from basic research to market”, project QSolid (Grant No. 13N16149).

[paper] SPICE Modeling of a Radiation Sensor

Miloš Marjanović 1, Stefan D. Ilić 2, Sandra Veljković 1, Nikola Mitrović 1, Umutcan Gurer 3, Ozan Yilmaz 3, Aysegul Kahraman 4, Aliekber Aktag 3, Huseyin Karacali 3, Erhan Budak 3, Danijel Danković 1, Goran Ristić 1 and Ercan Yilmaz 3

The SPICE Modeling of a Radiation Sensor Based on a MOSFET

with a Dielectric HfO2/SiO2 Double-Layer

Sensors 2025, 25(2), 546; DOI:10.3390/s25020546

1 Department of Microelectronics, Faculty of Electronic Engineering, University of Niš, Serbia
2 Center of Microelectronic Technologies, Institute of Chemistry, Technology and Metallurgy, University of Belgrade, Serbia
3 Faculty of Arts and Sciences, Bolu Abant Izzet Baysal University, Turkey
4 Department of Physics, Faculty of Arts and Sciences, Bursa Uludag University, Turkey

Abstract: We report on a procedure for extracting the SPICE model parameters of a RADFET sensor with a dielectric HfO2/SiO2 double-layer. RADFETs, traditionally fabricated as PMOS transistors with SiO2, are enhanced by incorporating high-k dielectric materials such as HfO2 to reduce oxide thickness in modern radiation sensors. The fabrication steps of the sensor are outlined, and model parameters, including the threshold voltage and transconductance, are extracted based on experimental data. Experimental setups for measuring electrical characteristics and irradiation are described, and a method for determining model parameters dependent on the accumulated dose is provided. A SPICE model card is proposed, including parameters for two dielectric thicknesses: (30/10) nm and (40/5) nm. The sensitivities of the sensors are 1.685mV/Gy and 0.78mV/Gy, respectively. The model is calibrated for doses up to 20Gy, and good agreement between experimental and simulation results validates the proposed model.

FIG: (a) Block diagram of the radiation source setup; 

(b) radiation setup in the TENMAK lab.

The corresponding SPICE model card is presented below:

.MODEL RADFET PMOS VTO={if(TYPE==1,-0.493-(1.54e-3*DOSE),-0.65433-(7.54E-4*DOSE))}

+KP={if(TYPE==1,8.897e-6-(1.493e-8*DOSE),1.14E-5-(2.511E-9*DOSE))} L=50e-6 W=600e-6

+TPG=0 LAMBDA={if(TYPE==1,3.901E-2-(2.165E-4*DOSE),2.0115E-2-(1.8575E-4*DOSE))}

Acknowledgements: This research was funded by North Atlantic Treaty Organization (NATO) SPS MYP under grant number G5974, by the project “High-k Dielectric RADFET for Detection of RN Treats”, and supported by the Ministry of Science, Technological Development and Innovation of the Republic of Serbia [grant number 451-03-65/2024-03/200102 and grant number 451-03-66/2024-03/200026].

[C4P] International Compact Modeling Conference

STRENGTHENING MODELING COLLABORATION WITH THE SEMICONDUCTOR INDUSTRY

International Compact Modeling Conference (ICMC 2025)

June 26-27, 2025; The Clift Royal Sonesta, San Francisco

IMPORTANT DATES Abstract Submission Deadline January 15, 2025 Acceptance Notifications March 10, 2025 Full Paper Submission Deadline April 20, 2025 ORGANIZING COMMITTEE General Chair Peter M. Lee Micron  Vice Chair Shahed Reza Sandia Lab Technical Program Chair Colin Shaw Silvaco Technical Program Vice Chair Gert-Jan Smit NXP  Treasurer Leigh Anne Clevenger Si2 Secretariat Conference Catalysts icmc@conferencecatalysts.com

The Compact Model Coalition (CMC) brings academia and industry partners together in the development and standardization of compact models for semiconductor devices. For 30 years now, the CMC has been instrumental in creating standardized and verified models for designers to use in their increasingly complex circuits for SPICE simulation. The CMC is organizing a new and innovative International Compact Modeling Conference. Cosponsored by IEEE EDS, it will focus uniquely on compact device models, their development and broad application in the semiconductor industry. You are invited to participate in the evolution of these models, guide model development to help circuit designers create the best circuit performance possible, and enable foundries to leverage the strength of their device fabrication to full extent. Join the world experts in design, process technology, and model development to discuss state-of-the-art semiconductor device modeling for a two-day in-person event in one location, offering a great opportunity to present and learn about this core element of circuit design and how to get the most from these global collaborations. We are seeking papers for oral or poster presentations in the following areas: APPLICATION OF DEVICE MODELS Innovative application of CMC standard device models Best practices, novel use, and benefits of standard device models in circuit design Use of compact models to demonstrate foundry device capabilities DEVICE MODEL DEVELOPMENT Modeling of physical phenomena: Statistical variation, reliability and aging, noise and fluctuations, high frequency effects, Electrostatic Discharge (ESD), self heating, layout effects, etc. Methodologies to assist in model development, practices for coding, quality assurance, circuit simulator integration, etc. Parameter extraction, measurement techniques, model calibration, validation, and verification methodologies, including solutions based on AI or Machine Learning. MODEL ENHANCEMENTS AND IMPLEMENTATIONS Model extensions to capture additional device features (leakage, noise, capacitance, second-order dependencies, …) or expand the operating range of existing devices (bias, power, temperature, frequency, etc.) Model enhancements to support the design of new or demanding circuits Model workflow, implementation, and integration into the design environment (PDK) Computing/simulation platforms, simulation algorithms, and methodologies to improve simulation performance (parallel processing, etc.) Models for established device types that currently lack standardization. MODELING FOR FUTURE/EMERGING TECHNOLOGIES AND APPLICATIONS Models for emerging device types or architectures on the horizon, such as, ferroelectric devices, silicon photonics, cryogenic, quantum computing, etc. Modeling of new physical phenomena in support of current and novel device technologies Novel device technologies currently being researched that could further revolutionize circuit performance, have implications in the design flow, and may become mainstream in the future Please submit your paper proposals in the form of a 2-page abstract for review by January 15, 2025 here 2025.si2-icmc.org. Acceptance notifications will be sent by March 10, 2025. Accepted contributions (for both oral and poster presentations) are expected to submit a camera-ready 4-page draft version of their papers by April 20, 2025 and final version by May 23, 2025 for publication in IEEE Xplore®.

[Program Highlights] 17th International MOS-AK Workshop Silicon Valley, December 11, 2024

 

17th International MOS-AK Workshop

Silicon Valley, December 11, 2024

Final MOS-AK Workshop Program

The 17th International MOS-AK Workshop on Compact/SPICE Modeling will online on Dec.11, 2024, in the timeframe of IEDM and Q4 CMC Meetings. This event is coorganized by Keysight Technologies, our local online host and partner, and the Extended MOS-AK TPC Committee. We cordially invite you to participate in the upcoming MOS-AK workshop, where you will have the opportunity to learn from leading experts in the field of the SPICE and Verilog-A modeling, OpenPDKs, and FOSS CAD/EDA IC designs. This event promises to be an invaluable experience for professionals and enthusiasts alike, offering deep insights and practical knowledge in these critical areas of the electron devices modeling and electronic design automation. The MOS-AK workshop program is available online and selected highlights are listed here:

 

Recent Compact Modeling Papers

[1] Hao Su, Yunfeng Xie, Yuhuan Lin, Haihan Wu, Wenxin Li, Zhizhao Ma, Yiyuan Cai, Xu Si, Shenghua Zhou Guangchong Hu, Yu He Feichi Zhou, Xiaoguang Liu, Longyang Lin, Yida Li, Hongyu Yu, and Kai Chen; "Characterizations and Framework Modeling of Bulk MOSFET Threshold Voltage Based on a Physical Charge-Based Model Down to 4 K." In 2024 IEEE European Solid-State Electronics Research Conference (ESSERC), pp. 733-736. IEEE, 2024. doi: 10.1109/ESSERC62670.2024.10719583

[2] Tung, Chien-Ting, Sayeef Salahuddin, and Chenming Hu; "A SPICE-Compatible Neural Network Compact Model for Efficient IC Simulations." In 2024 International Conference on Simulation of Semiconductor Processes and Devices (SISPAD), pp. 01-04. IEEE, 2024.

[3] Jana, Koustav, Shuhan Liu, Kasidit Toprasertpong, Qi Jiang, Sumaiya Wahid, Jimin Kang, Jian Chen, Eric Pop, and H-S. Philip Wong; "Modeling and Understanding Threshold Voltage and Subthreshold Swing in Ultrathin Channel Oxide Semiconductor Transistors." In 2024 International Conference on Simulation of Semiconductor Processes and Devices (SISPAD), pp. 01-04. IEEE, 2024.

[4] Manganaro, Gabriele. "Rethinking mixed-signal IC design." In 2024 IEEE European Solid-State Electronics Research Conference (ESSERC), pp. 552-556. IEEE, 2024

[5] Wager, John F., Jung Bae Kim, Daniel Severin, Zero Hung, Dong Kil Yim, Soo Young Choi, and Marcus Bender; "Dual-Layer Thin-Film Transistor Analysis and Design." IEEE Open Journal on Immersive Displays (2024).

[paper] FOSS support for CM with Verilog-A

Bűrmen, Árpád, Tadej Tuma, Iztok Fajfar, Janez Puhan, Žiga Rojec, Matevž Kunaver

and Sašo Tomažič

Free software support for compact modelling with Verilog-A

Informacije MIDEM 54, no. 4 (October 9, 2024)

<https://ojs.midem-drustvo.si/index.php/InfMIDEM/article/view/1999>

Abstract: Verilog-A is the analog subset of Verilog-AMS - a hardware description language for analog and mixed-signal systems. Verilog-A is commonly used for the distribution of compact models of semiconductor devices. For such models to be usable a Verilog-A compiler is required. The compiler converts the model equations into a form that can be used by the simulator. Such compilers have been supplied with commercial simulators for many years now. Free software alternatives are much more scarce and limited in the features they offer. The paper gives an overview of Verilog-A, Free software Verilog-A compilers, and Free software/Open source simulators that can simulate compact models defined in Verilog-A. Advantages and disadvantages of individual compilers and simulators are highlighted.

Tab: Comparison of Free software simulators

Asterisk denotes a feature under development as of Sep. 2024

Acknowledgements: This research was funded in part by the Slovenian Research Agency within the research program ICT4QoL—Information and Communications Technologies for Quality of Life, grant number P2-0246.

[paper] Compact Modeling of Hysteresis in OTFTs

Compact modeling of hysteresis in organic thin-film transistors

A. Romero^a, J.A. Jiménez-Tejada^a, R. Picos^b, D. Lara^a, J.B. Roldán^a, M.J. Deen^c

Organic Electronics 129 (2024) 107048

DOI : 10.1016/j.orgel.2024.107048

a Departamento de Electrónica y Tecnología de Computadores, CITIC-UGR, Uni Granada, Spain
b Department of Industrial Engineering and Construction, Universitat de les Illes Balears, Spain
c Department of Electrical and Computer Engineering, McMaster University, Canada

Abstract: In this work, we propose a model that describes the temporal evolution of the threshold voltage and trapped charge density in Thin-Film Transistors (TFTs) under dynamic conditions, paving the way for the characterization and modeling of memory transistors. The model is expressed as a first-order differential equation for the trapped charge density, which is controlled by a time constant and an independent term proportional to the drain current. The time-dependent threshold voltage is introduced in a previously developed compact model for TFTs with special consideration to the contact effects. The combination of both models and the use of an evolutionary parameter extraction procedure allow for reproducing the experimental dynamic behavior of TFTs. The results of the model and the evolutionary procedure have been validated with published experimental data of pentacene-based transistors. The procedure is able to simultaneously reproduce three kinds of experiments with different initialization routines and constraints in each of them: output and transfer characteristics with hysteresis and current transients characteristics

FIG: a.) Modeling the contact regions and intrinsic channel of an OTFT structure (a bottom contact configuration); b.)  Comparison of experimental transfer characteristics

Acknowledgements : The authors acknowledge support from the project PID2022 139586NB-44 funded by MCIN/AEI/10.13039/501100011033 and FEDER, EU. Funding for open access charge: Universidad de Granada / CBUA.

Appendix: Supplementary material related to this article can be found online.

[PhD] Transient Simulation of Frequency Domain Devices in Gnucap

Adding transient simulation of frequency domain devices to the Gnucap circuit simulator

Phd Thesis by Seán Higginbotham

Supervisor: Assistant Prof. Justin King

April 2024

Trinity College Dublin, The University of Dublin

College Green, Dublin 2, Ireland

Abstract: Radio frequency design constitutes a dominant element in the development of key communications technologies. Having accurate, robust, and widely accessible simulation methods is critical to ensuring continued advancements in this field, and guaranteeing the associated infrastructural and societal shifts that such technologies enable.

High frequency circuits invariably contain multiple non-linear components, which are naturally dealt with via time marching simulation of their time-domain analytic equations. However, including this alongside linear, generally dispersive, devices and effects, which are typically only characterised through a set of frequency-domain data describing the scattering response of an associated port-network, has traditionally been a problem for designers. Frequency-domain methods such as the harmonic balance technique and its successors have dominated radio frequency design for decades. However, such methods exhibit disadvantages in the context of modern circuits which are increasingly non-linear, and which operate with increasingly complicated modulated signals.

Various alternatives have been proposed, though as of yet no universally accepted method has emerged. Though harmonic balance will likely not be replaced, this project seeks to implement one such pure transient technique as an alternative. The proposed technique is based on using the vector-fitting algorithm to produce a model of the frequency response of the linear portnetwork, and then using a recursive convolution formulation to allow the time-domain response to be efficiently obtained from the port’s impulse response. An equivalent circuit companion model is developed from the resulting time-domain power-wave relation. This companion model allows the linear device to be directly included in a transient simulation alongside the analytic non-linear components, by way of providing a manner of computing the voltage and current on the network’s ports.

We implement the technique for one-port networks in a circuit driven by baseband signals. It is added to the free, open-source Gnucap circuit simulator as a ‘device plugin’. This report details how the implementation was done and provides results illustrating that it works as intended; the plugin can be installed by a user, who simply provides it with a file of frequency-domain data representing the port-network, and the plugin works naturally with the Gnucap transient solver to allow obtaining a transient solution of the overall circuit. A pure transient technique such as this does not require limiting assumptions or approximations on any components in the circuit and they are therefore preferable in certain contexts to frequency-domain methods like harmonic balance.

The project offers a significant contribution towards increasing the accessibility of radiofrequency electronics design and teaching.

 FIG: Summary of the traditional approach to simulating RF/MW circuits via HB, and the proposed pure transient approach implemented in this PhD Thesis

Acknowledgements: Seán Higginbotham would like to thank my M.A.I supervisor Dr. Justin King, whose previous work was the basis for this project. He provided invaluable insights and guidance which made the project both possible and an enjoyable experience, instilling curiosity at each discussion. Relevant academic references are included in the bibliography section. Acknowledgements of the dependancies used in the project code follow.

Gnucap is the creation of Albert Davis and is developed by him and others. It is provided under the GNU GPLv3, which is also the license that this project code is provided under on the associated GitHub repository.

See https://www.gnu.org/licenses/gpl-3.0.html. For the GNU GPLv3 license. Additionally, see the Gnucap repository here https://savannah.gnu.org/projects/ gnucap/.

LAPACK is a co-creation of The University of Tennessee and The University of Tennessee Research Foundation, The University of California Berkeley, and The University of Colorado Denver. See the user guide here https://netlib.org/lapack/.

The LAPACKE C bindings are the creation of Intel Corp.

The relevant licensing files are found within the source code and on the respective website.

Should the reader of this report have any questions or suggestions, please feel free to reach out at higginbs@tcd.ie, or via other channels such as the project GitHub located at https: //github.com/SHigginbotham/transient-sparam-gnucap. The project supervisor may also be of interest, available at justin.king@tcd.ie.

[paper] SiC Power MOSFET SPICE modelling

Akbar Ghulam

Accurate & Complete behaviourial SPICE modelling 

of commercial SiC Power MOSFET OF 1200V, 75A

25th EuroSimE, Catania, Italy, 2024, pp. 1-4,

DOI: 10.1109/EuroSimE60745.2024.10491420

* UNIPA Palermo (IT)

Abstract: Silicon Carbide (SiC) is proved to be an excellent replacement for Silicon in high voltage and high frequency applications due to its electro-thermal properties. Since SiC power MOSFETs have only recently been more widely available commercially, accurate simulation models are immediately required to forecast device behavior and facilitate circuit designs. The goal of this paper is to develop an accurate LTSPICE model based on a modified Enz-Krumenacher-Vittoz (EKV), MOSFET model for a 1200V, 30mΩ & 75ASiC power MOSFET “SCTW100N120G2AG” provided by STMicroelectronics that is currently on the market. The modified EKV model outperforms the reduced quadratic model by describing MOSFET behavior over different zones which are weak, moderate, and strong inversion zones with only a single equation. A wide range of experimental data was used to build the model's parameters. To estimate device performance in high frequency switching applications, the model has been expanded to include package parasitic components that include parasitic capacitances. The model's static and transient properties were simulated, and the results were compared with those acquired from the actual device.

FIG: The SiC MOSFET's circuit schematic utilizing a modified EKV model

Acknowledgements: We would like to thank STMicroelectronics, as for completion of this study has been greatly aided by their participation and availability of relevant data.

[paper] Symmetric BSIM-SOI

Chetan Kumar Dabhi, Dinesh Rajasekharan, Girish Pahwa, Debashish Nandi, Naveen Karumuri, Sreenidhi Turuvekere, Anupam Dutta, Balaji Swaminathan, Srikanth Srihari, Yogesh S. Chauhan, Sayeef Salahuddin, and Chenming Hu

Symmetric BSIM-SOI: A Compact Model for Dynamically Depleted SOI MOSFETs 

 in IEEE TED (2024)

Part I DOI: 10.1109/TED.2024.3363110

Part II DOI: 10.1109/TED.2024.3363117

1 Department of Electrical Engineering and Computer Sciences, UCB, CA, USA

2 Department of Electrical Engineering, IIT Kanpur, India

3 GlobalFoundries, Bengaluru, India

Abstract: In this article, we present a symmetric surface-potential-based model for dynamic depletion (DD) device operation of silicon-on-insulator (SOI) FETs for RF and analog IC design applications. The model accurately captures the device behavior in partial depletion (PD) and full depletion (FD) modes, as well as in the transition from PD to FD, based on device geometry, doping, and bias conditions. The model also exhibits an excellent source–drain symmetry during dc and small-signal simulations, resulting in error-free higher order harmonics. The model is fully scalable with bias, temperature, and geometry and has been validated extensively with real device data from the industry. The symmetric BSIM-SOI model is developed in Verilog-A and compatible with all commercial SPICE simulators.

FIG: (a) Schematic of a typical SOI MOSFET

(b) Cgg versus Vgb for different substrate bias, with the PD-to-FD transition 

Acknowledgment: The authors thanks the members of the Compact Model Coalition (CMC), particularly Geoffrey J. Coram and Jushan Xie, for testing the model and suggesting improvements. The authors appreciate the CMC QA team’s efforts in conducting a model quality check. Caixia Han and Xiao Sun from Cadence provided a few useful test cases. They thank Ananth Sundaram and Anamika Singh Pratiyush from GlobalFoundries India for the help and discussion regarding DDSOI model intricacies and development. Model code is available at BSIM Website <https://bsim.berkeley.edu/models/bsimsoi/>

[Open PDK] IEEE EDS DL at IISc Banglare

IEEE EDS/SSCS Bangalore Chapter Presents DL Series

FOSS TCAD/EDA Tools SPICE and Verilog-A

Modeling Flow Technology - Devices - Applications

W.Grabinski, MOS-AK (EU)

DATE AND TIME	LOCATION	HOSTS
Date: 07 Mar 2024 Time: 04:00 PM to 05:00 PM All times are (UTC+05:30) Chennai Add Event to Calendar iCal Google Calendar	Auditorium, Dept. of ESE, IISc Bangalore Karnataka India 560012	Bangalore Section Jt. Chapter ED15/SSC37

[FOSSDEM 2024] Open PDK Initiative

FOSDEM 2024 was a two-day event organized by volunteers to promote the widespread use of free and open source software. Took place at the ULB Solbosch campus in the beautiful city of Brussels (Belgium), FOSDEM is widely recognized as the best FOSS conference in Europe.

There were two DevRooms to discuss the status and further FOSS CAD/EDA IC design tools developments and open PDK initiative:

• Libre-SOC, FPGA and VLSI DevRoom
• Verilog-AMS in Gnucap
• FOSS CAD/EDA tools supporting the open access PDK initiative
• Open Hardware and CAD/CAM DevRoom
• ngspice circuit simulator - stand-alone and embedded into KiCad

FOSDEM'24 Inauguration Session

[paper] Neural Compact Modeling Framework

Eom, Seungjoon, Hyeok Yun, Hyundong Jang, Kyeongrae Cho, Seunghwan Lee, Jinsu Jeong, and Rock‐Hyun Baek

Neural Compact Modeling Framework for Flexible Model Parameter Selection with High Accuracy and Fast SPICE Simulation

Advanced Intelligent Systems (2023): 2300435

DOI: 10.1002/aisy.202300435

Department of Electrical Engineering, Pohang University of Science and Technology, Pohang 37673 (KR)

Abstract: Neural compact models are proposed to simplify device-modeling processes without requiring domain expertise. However, the existing models have certain limitations. Specifically, some models are not parameterized, while others compromise accuracy and speed, which limits their usefulness in multi-device applications and reduces the quality of circuit simulations. To address these drawbacks, a neural compact modeling framework with a flexible selection of technology-based model parameters using a two-stage neural network (NN) architecture is proposed. The proposed neural compact model comprises two NN components: one utilizes model parameters to program the other, which can then describe the current–voltage (IV) characteristics of the device. Unlike previous neural compact models, this two-stage network structure enables high accuracy and fast simulation program with integrated circuit emphasis (SPICE) simulation without any trade-off. The IV characteristics of 1000 amorphous indium–gallium–zinc-oxide thin-film transistor devices with different properties obtained through fully calibrated technology computer-aided design simulations are utilized to train and test the model and a highly precise neural compact model with an average IDS error of 0.27% and R2 DC characteristic values above 0.995 is acquired. Moreover, the proposed framework outperforms the previous neural compact modeling methods in terms of SPICE simulation speed, training speed, and accuracy.

Fig: a) The structure of a-IGZO TFT structure simulated with TCAD

b) Calibrated a-IGZO sub-gap DOS

Acknowledgements: This work was supported in part by the LG Display Company, in part by the Brain Korea 21 Fostering Outstanding Universities for Research (BK21 FOUR) program, in part by Institute of Information and Communications Technology Planning and Evaluation (IITP) grant funded by the Korea government (MSIT) (grant no. 2019-0-01906, Artificial Intelligence Graduate School Program [POSTECH]), in part by the Ministry of Trade, Industry and Energy (MOTIE) under grant no. 20020265, in part by Korea Semiconductor Research Consortium (KSRC) support program for the development of the future semiconductor device, and in part by the Technology Innovation Program (grant no. RS2023-00231985) funded by the Ministry of Trade, Industry and Energy (MOTIE, Korea) (grant no. 1415187390).

[paper] PSP RF Model

Xiaonian Liu1, 2, and Yansen Liu1, 2

Scalable PSP RF Model for 0.11 µm MOSFETs

Progress In Electromagnetics Research Letters, Vol. 113, 43–51, 2023

1 School of Physics and Electronics, Hunan Normal University, Changsha 410081, China
2 Key Laboratory of Physics and Devices in Post-Moore Era, College of Hunan Province, Changsha 410081, China.

Abstract: An accurate, efficient and scalable SPICE model is essential for modern integrated circuits design, especially for radio frequency (RF) circuit design. A PSP based scalable RF model is extracted and verified in 0.11 µm CMOS manufacturing process. The S parameter measurement system and open-short de-embedding technique is applied. The macro-model equivalent subcircuit and parameters extraction strategy are discussed. The extracted model can match the de-embedded S parameters data well. By combining the model parameters’ dependencies on each geometry quantity, the scalable expression of parameters with all geometry quantities included can be obtained. This work can be a reference for the RF MOSFETs modeling and RF circuit design.

Fig: The PSP RF subcircuit model and its S-par s fitting

results of NMOS with Wf = 2 µm, Lf = 0.12 µm, nf = 16

Acknowledgment: This work is supported by the National Natural Science Foundation of China under Grant 62204083, and the Youth Fund of Education Department of Hunan Province under Grant 21B0057.

[paper] PSP RF Model

Xiaonian Liu^{1, 2} and Yansen Liu¹

A Scalable PSP RF Model for 0.11 µm MOSFETs

Progress In Electromagnetics Research Letters, Vol. 113, 43–51, 2023

DOI :10.2528/PIERL23081405

1 School of Physics and Electronics, Hunan Normal University, Changsha 410081, China.
2 Key Laboratory of Physics and Devices in Post-Moore Era, College of Hunan Province, Changsha 410081, China.

Abstract : An accurate, efficient and scalable SPICE model is essential for modern integrated circuits design, especially for radio frequency (RF) circuit design. A PSP based scalable RF model is extracted and verified in 0.11 CMOS manufacturing process. The S parameter measurement system and open-short de-embedding technique is applied. The macro-model equivalent subcircuit and parameters extraction strategy are discussed. The extracted model can match the de-embedded S parameters data well. By combining the model parameters’ dependencies on each geometry quantity, the scalable expression of parameters with all geometry quantities included can be obtained. This work can be a reference for the RF MOSFETs modeling and RF circuit design.

Fig: The RF PSP Model Subcircuit

Acknowledgment : This work is supported by the National Natural Science Foundation of China under Grant 62204083, and the Youth Fund of Education Department of Hunan Province under Grant 21B0057.

[paper] Surface-Potential-Based Compact Modeling

M. Miura-Mattausch, T. Iizuka, H. Kikuchihara, H. J. Mattausch, and S. Saha

Evolution of Surface-Potential-Based Compact Modeling

IEEE EDS NEWSLETTER

OCTOBER 2023 VOL. 30, NO. 4 ISSN: 1074 1879

Abstract: Conventionally, a compact model of an electronic device is developed for utilization in circuit simulation. This means that the main task of the compact model is to accurately describe the characteristics of a device as a function of the applied voltages by simple equations in order to predict the performance of circuits using this device with sufficient precision. This overview article focuses on the compact modeling of the metal-oxide-semiconductor field-effect transistor (MOSFET)-device structure, which has the largest variety of applications. However, the modeling methodology is valid for any type of transistor or electronic device. The development of the compact modeling approach, based on the potential distribution induced within a transistor, is reviewed. The purpose of a compact model is to describe the transistor characteristics in a simple but accurate way, to enable correct circuit-performance prediction. Therefore, the basic physics of observed phenomena must be modeled by simplified and yet physically correct equations. To meet such requirements, potential-based modeling is a natural fit. A compact model and TCAD are both based on the same transistor equations. The difference is that TCAD considers the distribution of all physical quantities within a device, and a compact model integrates these distributions to calculate transistor characteristics at its nodes. The shortcomings of resulting simplifications, introduced for analytical integration, can be examined using TCAD, to identify observed phenomena still missing in the compact modeling. In this way, compact modeling is performed by learning from measurements macroscopically and from TCAD microscopically.

Fig: Schematic of a HV LDMOS FET (top) 

and its potential distribution (bottom)

[book] Microelectronic Circuits

Sedra, Adel S., Smith, Kenneth Carless, Carusone, 

Tony Chan, Gaudet, Vincent. 

Microelectronic Circuits. 

United Kingdom: Oxford University Press, 2020

Circuits by Sedra and Smith has served generations of electrical and computer engineering students as the best and most widely-used text for this required course. Respected equally as a textbook and reference, "Sedra/Smith" combines a thorough presentation of fundamentals with an introduction to present-day IC technology. It remains the best text for helping students progress from circuit analysis to circuit design, developing design skills and insights that are essential to successful practice in the field. Significantly revised with the input of two new coauthors, slimmed down, and updated with the latest innovations, Microelectronic Circuits, Eighth Edition, remains the gold standard in providing the most comprehensive, flexible, accurate, and design-oriented treatment of electronic circuits available today.

Appendix

• B. SPICE Device Models and Design with Simulation Examples

Model files for representative CMOS technologies are provided below:

• 0.8 um CMOS
• 0.35 um CMOS
• 0.18 um CMOS
• 45 nm CMOS
• 7nm FinFET

 

[11k online viewers] 7th Sino MOS-AK/Nanjing

Arbeitskreis Modellierung von Systemen und Parameterextraktion

Modeling of Systems and Parameter Extraction Working Group

7th Sino MOS-AK Workshop in Nanjing (CN)

August 11-13, 2023 (online/onsite)

Recent, consecutive, 7th Sino MOS-AK/Nanjing Workshop discussing the Compact/SPICE modeling and its Verilog-A Standardization reached 11k online viewers. The MOS-AK participants and online attendees have followed one day SiC-related device modeling training on August 11 featured presentations by experts currently working at Robert Bosch GmbH and then two days workshop with 24 R&D Compact/SPICE modeling presentations: