ISSCC 2019

2019

181 篇论文 · AI / ML (24) · Medical & Bio (21) · Wireline I/O (17) · Wireless (14) · Data Converters (14)

ISSCC 2019 Session 5 Image Sensors
A 256×256 40nm/90nm CMOS 3D-Stacked 120dBDynamic-Range Reconfigurable Time-Resolved SPAD Imager
Robert K. Henderson1, Nick Johnston1, Sam W. Hutchings1,
Kingdom 3 Heriot-Watt University, Edinburgh, United Kingdom 1 2 Light Detection and Ranging (LIDAR) applications pose extremely challenging dynamic range (DR) requirements on optical time-of-flight (ToF) receivers due to
Istvan Gyongy1, Tarek Al Abbas1, Neale Dutton2, Max Tyler3, Susan Chan
ISSCC 2019 Session 5 Image Sensors
A 32×32-Pixel 0.9THz Imager with Pixel-Parallel 12b VCO-Based ADC in 0.18μm CMOS
Sayuri Yokoyama1, Masayuki Ikebe1, Yuri Kanazawa1, Takahiro Ikegami1,
characteristics of signals in the terahertz band (100GHz to 10THz) located between the millimeter wave band and the infrared band. In particular, terahertz waves have higher spatial resolution than mm-waves. Moreover, th
Prasoon Ambalathankandy1, Shota Hiramatsu1, Eiichi Sano1, Yuma Takida2
ISSCC 2019 Session 6 Wireline I/O
A 100Gb/s 1.1pJ/b PAM-4 RX with Dual-Mode 1-Tap PAM4 / 3-Tap NRZ Speculative DFE in 14nm CMOS FinFET
Alessandro Cevrero, Ilter Ozkaya, Pier Andrea Francese, Matthias Brandli,
wireline communications has led to emerging standards in the 100Gb/s+ range using PAM-4 signaling. ADC-based receivers have demonstrated robust operation over channels with high losses (>20dB) [1], but their power consum
Christian Menolfi, Thomas Morf, Marcel Kossel, Lukas Kull, Danny Luu, M
ISSCC 2019 Session 6 Wireline I/O
A 60Gb/s PAM-4 ADC-DSP Transceiver in 7nm CMOS with SNR-Based Adaptive Power Scaling Achieving 6.9pJ/b at 32dB Loss
Marc-Andre LaCroix, Henry Wong, Yun Hua Liu, Huong Ho,
Semyon Lebedev, Petar Krotnev, Dorin Alexandru Nicolescu, Dmitry Petrov, Carlos Carvalho, Stephen Alie, Euhan Chong, Faisal Ahmed Musa, Davide Tonietto Huawei Technologies, Ottawa, ON, Canada With the introduction of PAM
ISSCC 2019 Session 6 Wireline I/O
A Sub-250mW 1-to-56Gb/s Continuous-Range PAM-4 42.5dB IL ADC/DAC-Based Transceiver in 7nm FinFET
Matteo Pisati, Fernando De Bernardinis, Paolo Pascale, Claudio Nani,
Marco Sosio, Enrico Pozzati, Nicola Ghittori, Federico Magni, Marco Garampazzi, Giacomino Bollati, Antonio Milani, Alberto Minuti, Fabio Giunco, Paola Uggetti, Ivan Fabiano, Nicola Codega, Alessandro Bosi, Nicola Carta,
ISSCC 2019 Session 6 Wireline I/O
A 180mW 56Gb/s DSP-Based Transceiver for HighDensity IOs in Data Center Switches in 7nm FinFET Technology
Tamer Ali1, Ramy Yousry*1, Henry Park*1, Ehung Chen*1,
Po-Shuan Weng2, Yi-Chieh Huang2, Chun-Cheng Liu2, Chien-Hua Wu2, Shih-Hao Huang2, Chungshi Lin2, Ke-Chung Wu2, Kun-Hung Tsai2, Kai-Wen Tan2, Ahmed ElShater1, Kuang-Ren Chen2, Wei-Hao Tsai2, Huan-Sheng Chen2, Weiyu Leng1,
ISSCC 2019 Session 6 Wireline I/O
A 400Gb/s Transceiver for PAM-4 Optical Direct-Detect Application in 16nm FinFET
C. Loi1, A. Mellati2, A. Tan3, A. Farhoodfar3, A. Tiruvur3, B. Helal3, B. Killips4,
F. Rad3, J. Riani3, J. Pernillo3, J. Sun1, J. Wong3, K. Abdelhalim2, K. Gopalakrishnan3, K. Kim2, L. Tse3, M. Davoodi2, M. Le2, M. Zhang3, M. Talegaonkar2, P. Prabha2, R. Mohanavelu3, S. Chong1, S. Forey4, S. Netto3, S.
ISSCC 2019 Session 6 Wireline I/O
A 128Gb/s 1.3pJ/b PAM-4 Transmitter with Reconfigurable 3-Tap FFE in 14nm CMOS
Zeynep Toprak-Deniz, Jonathan E. Proesel, John F. Bulzacchelli,
Herschel A. Ainspan, Timothy O. Dickson, Michael P. Beakes, Mounir Meghelli IBM T. J. Watson Research Center, Yorktown Heights, NY The ever-increasing demand for higher bandwidth continues to fuel the need for faster and
ISSCC 2019 Session 6 Wireline I/O
A 112Gb/s PAM-4 Voltage-Mode Transmitter with 4-Tap Two-Step FFE and Automatic Phase Alignment Techniques in 40nm CMOS
Pen-Jui Peng1, Yan-Ting Chen1, Sheng-Tsung Lai1, Chao-Hsuan Chen1,
transmitters to operate at higher speeds. The applications of 400GbE also push the transmitter to be designed at 112Gb/s for a single lane [1-2]. However, the use of advanced processes (<16nm) hardly reduces the costs. T
Hsiang-En Huang1, Ted Shih2 Yuan Ze University, Taoyuan City, Taiwan T
ISSCC 2019 Session 6 Wireline I/O
A 36Gb/s Adaptive Baud-Rate CDR with CTLE and 1-Tap DFE in 28nm CMOS
Danny Yoo1, Mohammad Bagherbeik1, Wahid Rahman1,
circuits (CDR) are ubiquitous in recent receiver designs as a means of lowering power consumption by sampling the data only once per UI. To further reduce power, prior works in pattern-based baud-rate PD
Ali Sheikholeslami1, Hirotaka Tamura2, Takayuki Shibasaki2 University
ISSCC 2019 Session 7 AI / ML
An 11.5TOPS/W 1024-MAC Butterfly Structure Dual-Core Sparsity-Aware Neural Processing Unit in 8nm Flagship Mobile SoC
Jinook Song1, Yunkyo Cho1, Jun-Seok Park1, Jun-Woo Jang2,
widely applied for image and speech recognition. Response time, connectivity, privacy and security drive applications towards mobile platforms rather than cloud. For mobile systems-on-a-chip (SoCs), energyefficient neural
Sehwan Lee2, Joon-Ho Song2, Jae-Gon Lee1, Inyup Kang1 Samsung Electron
ISSCC 2019 Session 7 AI / ML
A 20.5TOPS and 217.3GOPS/mm2 Multicore SoC with DNN Accelerator and Image Signal Processor Complying with ISO26262 for Automotive Applications
Yutaka Yamada, Toru Sano, Yasuki Tanabe, Yutaro Ishigaki,
Soichiro Hosoda, Fumihiko Hyuga, Akira Moriya, Ryuji Hada, Atsushi Masuda, Masato Uchiyama, Tomohiro Koizumi, Takanori Tamai, Nobuhiro Sato, Jun Tanabe, Katsuyuki Kimura, Ryusuke Murakami, Takashi Yoshikawa Toshiba Elect
ISSCC 2019 Session 7 AI / ML
An 879GOPS 243mW 80fps VGA Fully Visual CNN-SLAM Processor for Wide-Range Autonomous Exploration
Ziyun Li, Yu Chen, Luyao Gong, Lu Liu, Dennis Sylvester, David Blaauw, Hun-Seok Kim
University of Michigan, Ann Arbor, MI Simultaneous localization and mapping (SLAM) estimates an agent’s trajectory for all six degrees of freedom (6 DoF) and constructs a 3D map of an unknown surrounding. It is a fundame
ISSCC 2019 Session 7 AI / ML
A 2.1TFLOPS/W Mobile Deep RL Accelerator with Transposable PE Array and Experience Compression
Changhyeon Kim, Sanghoon Kang, Dongjoo Shin, Sungpill Choi,
but also for action control, so that an autonomous system, such as the robot, can perform human-like behaviors and operations. Unlike recognition tasks, real-time operation is important in action control, and it is too s
Youngwoo Kim, Hoi-Jun Yoo KAIST, Daejeon, Korea Recently, deep neural
ISSCC 2019 Session 7 AI / ML
A 65nm 0.39-to-140.3TOPS/W 1-to-12b Unified NeuralNetwork Processor Using Block-Circulant-Enabled Transpose-Domain Acceleration with 8.1× Higher TOPS/mm2 and 6T HBST-TRAM-Based 2D Data-Reuse Architecture
Jinshan Yue1, Ruoyang Liu1, Wenyu Sun1, Zhe Yuan1, Zhibo Wang1,
Yung-Ning Tu2, Yi-Ju Chen2, Ao Ren3, Yanzhi Wang3, Meng-Fan Chang2, Xueqing Li1, Huazhong Yang1, Yongpan Liu1 Tsinghua University, Beijing, China National Tsing Hua University, Hsinchu, Taiwan 3 Northeastern University,
ISSCC 2019 Session 7 AI / ML
A 65nm 236.5nJ/Classification Neuromorphic Processor with 7.5% Energy Overhead On-Chip Learning Using Direct Spike-Only Feedback
Seoul National University, Seoul, Korea, spikes, greatly reducing computation and global interconnects. While the single
classifying handwritten digits, the learning rule can be directly adopted in general fully-connected networks with different network structures and hence can be extended to other applications. For example, the algorithm
ISSCC 2019 Session 7 AI / ML
LNPU: A 25.3TFLOPS/W Sparse Deep-Neural-Network Learning Processor with Fine-Grained Mixed Precision of FP8-FP16
Jinsu Lee, Juhyoung Lee, Donghyeon Han, Jinmook Lee, Gwangtae Park, Hoi-Jun Yoo
for energy-efficient deep learning (DL) acceleration [1-6]. Most prior DNN inference accelerators are trained in the cloud using public datasets; parameters are then downloaded to implement AI [1-5]. However, local DNN le
KAIST, Daejeon, Korea Recently, deep neural network (DNN) hardware acc
ISSCC 2019 Session 8 Power Management
A Continuous-Input-Current Passive-Stacked Third-Order Buck Converter Achieving 0.7W/mm2 Power Density and 94% Peak Efficiency
Abdullah Abdulslam, Patrick P. Mercier
The power density and efficiency of power-management integrated circuits (PMICs) is playing an increasingly important role in the miniaturization of modern computing platforms. Small inductors can be used to help miniatur
University of California, San Diego, La Jolla, CA
ISSCC 2019 Session 8 Power Management
A 10.9W 93.4%-Efficient (27W 97%-Efficient) FlyingInductor Hybrid DC-DC Converter Suitable for 1-Cell (2-Cell) Battery Charging Applications
Casey Hardy, Hanh-Phuc Le
The annual increase in performance and feature density of mobile products has driven the need for batteries with higher capacity and thus higher power-delivery solutions to maintain sensible charging times. The USB-C pow
University of Colorado, Boulder, CO
ISSCC 2019 Session 8 Power Management
Fully Integrated Buck Converter with 78% Efficiency at 365mW Output Power Enabled by Switched-InductorCapacitor Topology and Inductor Current Reduction Technique
Nghia Tang1, Bai Nguyen1, Yangyang Tang2, Wookpyo Hong1,
power consumption of system-on-chip by providing point-of-load regulation with dynamic voltage scaling [1]. As a core component of a buck converter, an inductor with large inductance and small resistance is desirable for
Zhiyuan Zhou1, Deukhyoun Heo1 Washington State University, Pullman, WA
ISSCC 2019 Session 8 Power Management
A Fully Integrated Voltage Regulator in 14nm CMOS with
Package-Embedded Air-Core Inductor Featuring SelfTrimmed, Digitally Controlled Variable On-Time, Discontinuous Conductio
Christopher Schaef1, Nachiket Desai1, Harish Krishnamurthy1, Sheldon Weng1, Huong Do2, William Lambert2, Kaladhar Radhakrishnan2, Krishnan Ravichandran1, James Tschanz1, Vivek De1 Intel, Hillsboro, OR Intel, Chandler, AZ
ISSCC 2019 Session 8 Power Management
A Fully Integrated 85%-Peak-Efficiency Hybrid MultiRatio Resonant DC-DC Converter with 3.0-to-4.5V Input and 500µA-to-120mA Load Range by a fast capacitive level shifter, which results in very energy efficient highfrequency signal shifting. In contrast, pulsed cascode level shifters are used for
SwCR, since the controller sets the switch conductance information Sw<0:7> at, a rate, which is far below fsw. These lev
static losses compared to the fast dynamic level shifters. Peter Renz1, Maik Kaufmann1, Michael Lueders2, Bernhard Wicht1 Leibniz University Hannover, Hannover, Germany 2 Texas Instruments, Freising, Germany 1 DC-DC conv
ISSCC 2019 Session 8 Power Management
A 2MHz 4-to-60VIN Buck-Boost Converter for Automotive Use Achieving 95% Efficiency and CISPR 25 Class 5 Standard
Jing Xue, Min Kyu Song, Xugang Ke, Min Chen, Leonard Shtargot
High-voltage highly efficient switching power converters have gained high popularity in automotive applications. Powered by the car battery (VIN), the converters need to experience a wide VIN range from 4V to 60V under ha
Analog Devices, Milpitas, CA
ISSCC 2019 Session 9 mm-Wave
Toward Automotive Surround-View Radars
Chih-Ming Hung1, Alex TC Lin1, BC Peng1, Hua Wang2, Jui-Lin Hsu1, Yen-Ju Lu1,
Weishow Hsu1, Jing-Hong Conan Zhan1, Brian Juan1, Chi-Hang Lok1, Sam Lee1, PC Hsiao1, Qiang Zhou2, Mark Wei1, Hsiang-Yun Chu1, Yu-Lun Chen1, Chao-Ching Hung1, Kevin Fong1, Po-Chun Huang1, Pierce Chen1, Sheng-Yuan Su1, Ya
ISSCC 2019 Session 9 mm-Wave
A 192-Virtual-Receiver 77/79GHz GMSK Code-Domain MIMO Radar System-on-Chip
Vito Giannini, Marius Goldenberg, Aria Eshraghi, James Maligeorgos,
Lysander Lim, Ryan Lobo, Dave Welland, Chung-Kai Chow, Andrew Dornbusch, Tim Dupuis, Struan Vaz, Fred Rush, Paul Bassett, Hong Kim, Monier Maher, Otto Schmid, Curtis Davis, Manju Hegde Uhnder, Austin, TX MIMO radars tran
ISSCC 2019 Session 9 mm-Wave
A 680µW Burst-Chirp UWB Radar Transceiver for Vital Signs and Occupancy Sensing up to 15m Distance
Yao-Hong Liu1, Sunil Sheelavant1,2, Marco Mercuri1, Paul Mateman1,
Johan Dijkhuis1, Wilfried Zomagboguelou1,3, Arjan Breeschoten1, Stefano Traferro1, Yan Zhan1, Tom Torf4, Christian Bachmann1, Pieter Harpe3, Masoud Babaie2 imec - Netherlands, Eindhoven, The Netherlands Delft University
ISSCC 2019 Session 9 mm-Wave
A 145GHz FMCW-Radar Transceiver in 28nm CMOS
Akshay Visweswaran, Kristof Vaesen, Siddhartha Sinha, Ilja Ocket,
require a high range resolution. This paper presents a 145GHz FMCW radar transceiver with on-chip antennas in 28nm bulk CMOS. An RF bandwidth of 13GHz yields an 11mm range resolution, and the high RF carrier permits grea
Miguel Glassee, Claude Desset, Andre Bourdoux, Piet Wambacq imec, Leuv
ISSCC 2019 Session 9 mm-Wave
An 80Gb/s 300GHz-Band Single-Chip CMOS Transceiver
Sangyeop Lee1, Ruibing Dong1, Takeshi Yoshida1, Shuhei Amakawa1,
Technology, Koganei, Japan 3 Panasonic, Yokohama, Japan 1 2 IEEE Standard 802.15.3d, published in October 2017, defines a high-data-rate wireless physical layer that enables up to 100Gb/s using the lower THz frequency ran
Shinsuke Hara2, Akifumi Kasamatsu2, Junji Sato3, Minoru Fujishima1 Hir
ISSCC 2019 Session 9 mm-Wave
A 42.2Gb/s 4.3pJ/b 60GHz Digital Transmitter with 12b/Symbol Polarization MIMO
Chintan Thakkar, Stefan Shopov, Anandaroop Chakrabarti,
in modern-day handheld computing devices around the unlicensed 57-to-71GHz band, the IEEE 802.11ay standard [1] supports 2-to-4× bonding of 2.16GHz channels (i.e. 4.32 to 8.64GHz), constellations up to 64QAM, and spatial
Shuhei Yamada, Debabani Choudhury, James Jaussi, Bryan Casper Intel, H
ISSCC 2019 Session 9 mm-Wave
A Scalable 71-to-76GHz 64-Element Phased-Array Transceiver Module with 2×2 Direct-Conversion IC in 22nm FinFET CMOS Technology
Stefano Pellerano1, Steven Callender1, Woorim Shin1, Yanjie Wang2,
Somnath Kundu1, Abhishek Agrawal1, Peter Sagazio1, Brent Carlton1, Farhana Sheikh1, Arnaud Amadjikpe1, William Lambert3, Divya Shree Vemparala1, Mark Chakravorti1, Satoshi Suzuki1, Robert Flory1, Chris Hull1 Intel, Hills
ISSCC 2019 Session 9 mm-Wave
A 28GHz 20.3%-Transmitter-Efficiency 1.5°-Phase-Error Beamforming Front-End IC with Embedded Switches and Dual-Vector Variable-Gain Phase Shifters
Jinseok Park*1, Seungchan Lee*1, Dongho Lee2, Songcheol Hong1
Hanbat National University, Daejeon, Korea *Equally-Credited Authors (ECAs) 1 respectively. The NF of RX at 28GHz is 4.58dB, which is increased only by 1dB compared to simulated NF of LNA without switch and PA. NF variat
KAIST, Daejeon, Korea