Digital Processors

Sumanth Gururajarao1, Rolf Lagerquist1, Jin Son1, Gordon Gammie1, Gordon Lin2, Achuta Thippana1, Kent Li1, Manzur Rahman1, Wuan Kuo2, David Yen2, Yi-Chang Zhuang2, Ue Fu2, Hung-Wei Wang2, Mark Peng3, Cheng-Yuh Wu2, Taner

Kok Hong Chan1, Andy Tong1, Sean Atsatt1, Dana How1, Peter McElheny1, Keith Duwel1, Jeffrey Schulz1, Darren Faulkner3, Gopal Iyer1, George Chen1, Hee Kong Phoon4, Han Wooi Lim4, Wei-Yee Koay4, Ty Garibay3 Intel, San Jose

Shane Southard1, Hugh McIntyre3, Amy Novak1, Stephen Kosonocky2, Ravi Jotwani1, Alex Schaefer1, Edward Chang2, Joshua Bell1, Michael Co1 AMD, Austin, TX AMD, Fort Collins, CO 3 AMD, Sunnyvale, CA 1 2 Codenamed “Zen”, AMD

Rahul Rao3, Jose Paredes2, Michael Floyd2, Michael Sperling4, Ryan Kruse2, Vinod Ramadurai2, Ryan Nett2, Saiful Islam2, Juergen Pille5, Donald Plass4 IBM, Yorktown Heights, NY IBM, Austin, TX 3 IBM, Bangalore, India 4 IB

Pawel Owczarczyk3, Eric J. Fluhr1, Joshua Friedrich1, Paul Muench3, Timothy Diemoz3, Pierce Chuang2, Christos Vezyrtzis2 IBM, Austin, TX IBM, Yorktown Heights, NY 3 IBM, Poughkeepsie, NY 1 2 Increasing transistor counts

Clock distribution in modern SoCs consumes a significant fraction of total chip power. To reduce clock distribution power, resonant clocking schemes, where an inductive reactance is used to cancel the capacitive reactanc

Clock network optimization is substantially affected by the operating voltage VDD, as the clock skew is dominated by different mechanisms and has a different balance between wire and repeater delay at different VDD (Fig.

Richard Rizzolo3, Tobias Webel4, Thomas Strach4, Otto Torreiter4, Preetham Lobo5, Alper Buyuktosunoglu1, Ramon Bertran1, Michael Floyd6, Malcolm Ware6, Gerard Salem7, Sean Carey8, Phillip Restle1 IBM Research, Yorktown H

they replace Point-Doubling with an essentially free Frobenius Endomorphism [7]. Efficient micro-coding and a Serial Input Parallel Output (SIPO) finite field multiplier ensure that only 2 additional registers (t, z) are

machine learning (ML) and the internet-of-things (IoT) have resulted in a renewed interest in analog matrix-vector multiplication (MvM) accelerators [1-3]. Classification based tasks have exploited low-to-medium resoluti

increasingly intelligent and context-aware. Sound is an attractive sensory modality because it is information-rich but not as computationally demanding as alternatives such as vision. New applications of ultra-low power

Piezoelectric actuators are used in a growing range of applications, e.g., haptic feedback systems, cooling fans, and microrobots. However, to fully realize their potential, these actuators require drivers able to effici

Intel, Hillsboro, OR 1 2 Piezoelectric transducer (PT) is an emerging energy-based technology for electrosurgery. With proper driving signals, the PT is utilized as an electric scalpel as shown in Fig. 21.4.1. It convert

intraocular pressure (IOP) in the anterior chamber of the eye can damage the optic nerve causing irreversible blindness [1]. An IOP monitoring microsystem (IMM) implanted in the interior chamber of the eye is required to

ventilation because it is the only real-time lung imaging method without large equipment [1-2]. However, previous EIT systems just provided 2D cross-sectional image with limited spatial information of the lung and unnegl

emerging as a platform for small molecule detection in challenging environments [1]. A key barrier to widespread deployment of autonomous bacterial sensors is the detection of low-level bioluminescence, which is typicall

National Chiao Tung University, Hsinchu, Taiwan 1 2 DNA sequencing is the process of determining the precise order of nucleotides (A, C, G, T) within a DNA molecule and is now indispensable for genetics and medical resea

Qing Dong1, Yen-Po Chen1, Laura Fick1, Xun Sun1, Ron Dreslinski1, Trevor Mudge1, Hun Seok Kim1, David Blaauw1, Dennis Sylvester1 University of Michigan, Ann Arbor, MI CubeWorks, Ann Arbor, MI 1 2 Deep learning has proven

for the next-generation UI/UX of wearable devices. A FR system, shown in Fig. 14.6.1, was developed as a life-cycle analyzer or a personal black box, constantly recording the people we meet, along with time and place inf

Analog Devices, Cambridge, MA 1 Previous work such as [4] provided micropower VADs that can be used in quiet environments or in applications that tolerate false alarms. In our application, false alarms will unnecessarily

(IoT) devices with the capability to interpret the complex, noisy real-world data arising from sensorrich systems. Achieving sufficient energy efficiency to execute ML workloads on an edge-device necessitates specialized

Zhe Yuan1, Chien-Chen Lin2, Qi Wei1, Yu Wang1, Ya-Chin King2, Chrong-Jung Lin2, Pedram Khalili3, Kang-Lung Wang3, Meng-Fan Chang2, Huazhong Yang1 Tsinghua University, Beijing, China, National Tsing Hua University, Hsinch

many application areas, such as IoT and wearables. While for some applications, integer and fixed-point processor instructions suffice, others (e.g. simultaneous localization and mapping – SLAM, stereo vision, nonlinear

car information systems (commonly referred to as car infotainment) is expanding from dedicated navigation systems to joint car-cockpit systems, including the dashboard meter, telematics for the internet/cloud, and advanc

Chi Lan Phuong Nguyen2, Tetsuya Shibayama1, Kenichi Iwata1, Katsuya Mizumoto1, Takahiro Irita3, Hirotaka Hara3, Toshihiro Hattori1 Renesas System Design, Tokyo, Japan, Renesas Design Vietnam, Ho Chi Minh City, Vietnam, 3

C.J. Chung1, Sumanth Gururajarao1, Ping Kao2, Anand Rajagopalan1, Anirban Saha3, Amit Jain4, Ericbill Wang2, Shichin Ouyang5, Huajun Wen1, Achuta Thippana1, HsinChen Chen1, Syed Rahman1, Minh Chau1, Anshul Varma1, Brian

Ravinder Rachala1, Sriram Sambamurthy1, Steven Liepe2, Miguel Rodriguez2, Tom Burd3, Adam Clark4, Michael Austin1, Samuel Naffziger2 AMD, Austin, TX, AMD, Fort Collins, CO, 3 AMD, Sunnyvale, CA, 4 AMD, Markham, ON, Canad

Muhammad Abozaed, Yair Talker, Ziv Shmuely, Saher Abu Rahme transitions in the victim neighborhood circuits should not exceed a crosstalk limit during normal operation. Another parameter that must be factored is reliabil

Jim Farrell1, Dave Johnson2, Guhan Krishnan1, Hugh McIntyre3, Edward McLellan1, Samuel Naffziger2, Russell Schreiber4, Sriram Sundaram4, Jonathan White1 AMD, Boxborough, MA, 2AMD, Fort Collins, CO, 3 AMD, Sunnyvale, CA,

Droop, Temperature and Aging Jaydeep P. Kulkarni, Carlos Tokunaga, Paolo Aseron, Trang Nguyen Jr, Charles Augustine, James Tschanz, Vivek De Intel, Hillsboro, OR 8-transistor (8T) cell 1-read/1-write (1R1W) register file

Arvind Raghavan1, Charles Morganti2, Chris Houghton1, Dan Krueger2, Olivier Franza1, Jayen Desai2, Jason Crop2, Dave Bradley2, Chris Bostak2, Sal Bhimji1, Matt Becker1 Intel, Hudson, MA, 2Intel, Fort Collins, CO 1 The ne

Huy N. Nguyen2, Mihir Pandya2 IBM Research, Rüschlikon, Switzerland, Freescale Semiconductor, Austin, TX 1 2 MicroServers integrate an entire server motherboard into a single Server-on-aChip (SoC), excluding DRAM, NOR-bo

Curtis McAllister, Ha Pham, Sebastian Turullols, Jinuk Luke Shin, Yifan YangGong, Haowei Zhang Oracle, Redwood Shores, CA The goal of the power management system of Oracle’s SPARC M7 CPU [1] is to maximize the performanc

Francis Schumacher, Venkat Krishnaswamy, Hoyeol Cho, Sudesna Dash, Robert Masleid, Chaoyang Zheng, Yuanjung David Lin, Paul Loewenstein, Heechoul Park, Vijay Srinivasan, Dawei Huang, Changku Hwang, Wenjay Hsu, Curtis McA

Donald Plass2, Yuen Chan2, Sean Carey2, Gerard Salem4, Friedrich Schroeder5, Frank Malgioglio2, Guenter Mayer5, Christopher Berry2, Michael Wood2, Yiu-Hing Chan2, Mark Mayo2, John Isakson3, Charudhattan Nagarajan6, Tobia

Ichiro Lin1, HsinChen Chen1, Wuan Kuo2, Anand Rajagopalan1, Wei-Zheng Ge1, Rolf Lagerquist1, Syed Rahman1, CJ Chung1, Simon Wang2, Lee-Kee Wong2, Yi-Chang Zhuang2, Kent Li1, Jidong Wang1, Minh Chau1, Yijing Liu1, Daniel

essential in emerging computer vision applications. Although active sensing methods can provide an accurate indoor depth map, they have limited resolution and consume significant power, such as the 2.1W time-of-flight se

Kwangil Kim, Ken Shin, Yohan Kwon, Heungchul Oh, Jaeyoung Lim, Dong-wook Lee, Jongho Lee, Inpyo Hong, Kyungkuk Chae, Heon-Hee Lee, Sung-Wook Lee, Seongho Song, Chung-Hee Kim, Jin-Soo Park, Heesoo Kim, Sunghee Yun, Uk-Rae

The latest multiple-input multiple-output (MIMO) wireless systems have adopted iterative detection and decoding (IDD) to reduce the signal-to-noise ratio (SNR) required for a reliable transmission. An IDD system consists

Han-Liang Chou, Chih-Ming Wang, Tung-Hsing Wu, Hue-Min Lin, Yi-Hsin Huang, Chia-Yun Cheng, Ting-An Lin, Chun-Chia Chen, Yu-Kun Lin, Min-Hao Chiu, Wei-Cing Li, Sheng-Jen Wang, Yen-Chieh Lai, Ping Chao, Chih-Da Chien, Meng

University of California, Los Angeles, CA Compressive sensing (CS) is a promising solution for low-power on-body sensors for 24/7 wireless health monitoring [1]. In such an application, a mobile data aggregator performin

Embedded sensing systems conventionally perform A-to-D conversion followed by signal analysis. In many applications, the analysis of interest is inference (e.g., classification), but the sensor signals involved are too c

Microwatt object recognition is being considered for many applications, such as autonomous micro-air-vehicle (MAV) navigation, a vision-based wake-up or user authentication for the smartphones, and a gesture recognition-

Mayu Okumura, Manabu Nishiyama, Tadakazu Nomura, Kazushige Oma, Nobuhiro Sato, Moriyasu Banno, Hiroo Hayashi, Takashi Miyamori Toshiba, Kawasaki, Japan Image recognition technologies have gained prominence in a variety o

next-generation mainstream wearable devices. However, previous HMD systems [1] have had limited application, primarily due to their lacking a smart user interface (UI) and user experience (UX). Since HMD systems have a s

Seyed Arash Mirhaj1, Yen-Cheng Kuan1, Huan-Neng Chen2, Chewn-Pu Jou2, Ming-Hsien Tsai2, Fu-Lung Hsueh2, Mau-Chung Frank Chang1 University of California, Los Angeles, CA, 2TSMC, Hsinchu, Taiwan 1 The noise performance of

architectures, from a planar structure to a FinFET to achieve decreased leakage, further downscaling, and better sub-threshold slope, even under a lower power supply [1]. Downscaling trends have forced the analog semicon

SRAM register files have sensitive circuitry and often operate with high switching activity and at high temperature. This makes them particularly vulnerable to aging by negative-bias temperature instability (NBTI) degrad

ARM, Cambridge, United Kingdom The current trend for System-on-Chip (SoC) compute subsystems is to improve energy efficiency, while operating at a similar power budget as previous generations. Reduced supply voltages and

(ADPLLs) offer faster locking time, easier portability and better performance in advanced semiconductor processes as compared to analog PLLs. Advanced FinFET devices exhibit better gm and ION than planar devices [1], but