Yannis Tsividis
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Yannis Tsividis has a total of 22 co-authors. Here are the top co-authors by number of shared publications:
- Sharvil Patil
Analog Devices Canada, Toronto, ON, Canada
2 shared publications - Yu Chen
Xian Technological University,School of Electronic Information Engineering,XIAN,CHINA
2 shared publications Alin Ratiu
CEA-LETI-MINATEC, Grenoble, France
1 shared publication- Bob Schell
Department of Electrical Engineering, Columbia University, New York, NY 10027 U.S.A.
1 shared publication - Boris Murmann
University of Hawai‘i at Manoa
1 shared publication Christos Vezyrtzis
Yorktown Heights, NY, USA
1 shared publication
Latest Publications
- Open Access
- 5 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-6
SPADES: A 0.54-GFLOPS/W Sparse Matrix Vector Multiplication Accelerator Featuring On-the-Fly GZIP Decompression for 3.36X Reduction in Off-Chip Data Movement
We propose SPADES, the first sparse matrix vector multiplication (SpMV) accelerator incorporating online GZIP decompression. The goal of the decompression is to reduce off-chip data movement, which has become a major source of energy consumption in SpMV accelerators. Our proposed GZIP-SpMV flow achieves an average compression ratio of 3.36 for the sparse matrix CSC data, reducing the off-chip data traffic by 70%. We fabricated SPADES in TSMC 28-nm with a die area of 2.47 mm2. Compared with the prior SpMV accelerator, SPADES achieves 2.32X better energy efficiency.
- 5 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Open AccessSPADES: A 0.54-GFLOPS/W Sparse Matrix Vector Multiplication Accelerator Featuring On-the-Fly GZIP Decompression for 3.36X Reduction in Off-Chip Data Movement
We propose SPADES, the first sparse matrix vector multiplication (SpMV) accelerator incorporating online GZIP decompression. The goal of the decompression is to reduce off-chip data movement, which has become a major source of energy consumption in SpMV accelerators. Our proposed GZIP-SpMV flow achieves an average compression ratio of 3.36 for the sparse matrix CSC data, reducing the off-chip data traffic by 70%. We fabricated SPADES in TSMC 28-nm with a die area of 2.47 mm2. Compared with the prior SpMV accelerator, SPADES achieves 2.32X better energy efficiency.
- 3 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Solid-State Circuits Magazine
- Vol. 16 (2) , 84-88
T. R. Viswanathan (1937–2023) [Society News]
Dr. T. R. Viswanathan, fondly called “TR,” passed away in December 2023. He influenced several generations of students and colleagues throughout his career, which was spread across multiple countries, universities, and industries. TR returned to India after completing his Ph.D. degree at the University of Saskatchewan as a founding faculty member of the Indian Institute of Technology (IIT), Kanpur. After 15 years at IIT Kanpur, he migrated to Canada and then the United States, working in universities and industry. This article features reminiscences of TR from members of our community.
- 3 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Solid-State Circuits Magazine
- Vol. 16 (2) , 84-88
T. R. Viswanathan (1937–2023) [Society News]
Dr. T. R. Viswanathan, fondly called “TR,” passed away in December 2023. He influenced several generations of students and colleagues throughout his career, which was spread across multiple countries, universities, and industries. TR returned to India after completing his Ph.D. degree at the University of Saskatchewan as a founding faculty member of the Indian Institute of Technology (IIT), Kanpur. After 15 years at IIT Kanpur, he migrated to Canada and then the United States, working in universities and industry. This article features reminiscences of TR from members of our community.
- 13 November 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 66 (3) , 1017-1030
Signal Encoding and Processing in Continuous Time Using a Cascade of Digital Delays
We consider a digital encoding/processing system in which both the analog-to-digital converter and the digital signal processor are implemented by repeating a single fundamental design unit: the digital delay. Timing becomes an integral part of the resulting signal representation and processing, thereby promising to improve with technology scaling. System properties are studied and design considerations for a potential integrated implementation are discussed. Simulation results confirm the theory.
- 13 November 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 66 (3) , 1017-1030
Signal Encoding and Processing in Continuous Time Using a Cascade of Digital Delays
We consider a digital encoding/processing system in which both the analog-to-digital converter and the digital signal processor are implemented by repeating a single fundamental design unit: the digital delay. Timing becomes an integral part of the resulting signal representation and processing, thereby promising to improve with technology scaling. System properties are studied and design considerations for a potential integrated implementation are discussed. Simulation results confirm the theory.
- Open Access
- 4 December 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. 53 (2) , 418-430
A Continuous-Time Digital IIR Filter With Signal-Derived Timing and Fully Agile Power Consumption
Presented is the first continuous-time (CT) digital infinite impulse response (IIR) filter working on signal-derived timing in lieu of a clock. We introduce a novel design method which enables the design of high-order IIR filters using only two tap delays. An event-grouping technique is also introduced to prevent parasitic oscillations in the presence of tap delay mismatches. The 1.2-V, 65-nm CMOS prototype implements a sixth-order IIR filter, with a maximum input rate of 20 Msample/s and a stop-band rejection of more than 80 dB. Without using any power-down circuitry, the chip's power consumption tracks the input activity in a fully agile manner, and varies by more than 50$x $, from 0.04 to 2.32 mW. The filter achieves an figure of merit (FoM) which competes with that of discrete-time (DT) filters, while avoiding the use of a clock and an antialiasing filter. Compared to prior art in CT digital signal processings, the prototype achieves 45-dB improvement in stop-band rejection and 9x smaller delay line area. For the first time, the filtered CT digital signal is converted to synchronous mode at the end of signal chain, allowing integration with DT digital systems.
- 4 December 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. 53 (2) , 418-430
Open AccessA Continuous-Time Digital IIR Filter With Signal-Derived Timing and Fully Agile Power Consumption
Presented is the first continuous-time (CT) digital infinite impulse response (IIR) filter working on signal-derived timing in lieu of a clock. We introduce a novel design method which enables the design of high-order IIR filters using only two tap delays. An event-grouping technique is also introduced to prevent parasitic oscillations in the presence of tap delay mismatches. The 1.2-V, 65-nm CMOS prototype implements a sixth-order IIR filter, with a maximum input rate of 20 Msample/s and a stop-band rejection of more than 80 dB. Without using any power-down circuitry, the chip's power consumption tracks the input activity in a fully agile manner, and varies by more than 50$x $, from 0.04 to 2.32 mW. The filter achieves an figure of merit (FoM) which competes with that of discrete-time (DT) filters, while avoiding the use of a clock and an antialiasing filter. Compared to prior art in CT digital signal processings, the prototype achieves 45-dB improvement in stop-band rejection and 9x smaller delay line area. For the first time, the filtered CT digital signal is converted to synchronous mode at the end of signal chain, allowing integration with DT digital systems.
- 1 October 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1605-1609
Signal processing in continuous time using asynchronous techniques
This paper reviews previous work on continuous-time digital signal processing systems. The operations of these systems avoid the use of a clock and can closely track the input activity. The two major building blocks, continuous-time analog-to-digital converter (CT ADC) and continuous-time digital signal processor (CT DSP) are discussed in detail, with several examples presented.
- 1 October 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1605-1609
Signal processing in continuous time using asynchronous techniques
This paper reviews previous work on continuous-time digital signal processing systems. The operations of these systems avoid the use of a clock and can closely track the input activity. The two major building blocks, continuous-time analog-to-digital converter (CT ADC) and continuous-time digital signal processor (CT DSP) are discussed in detail, with several examples presented.
- 1 June 2015
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. C160-C161
A 310fJ/conv-step 0.0032mm2 error-shaping alias-free asynchronous ADC
We present a programmable, quantization error spectral shaping, alias-free asynchronous ADC suited for clockless, continuous-time DSP in receivers with modest SNDR requirements and a tight power budget. Implemented in 0.65V 28nm FDSOI, the 0.0032mm 2 ADC achieves 32dB-42dB SNDR over a 10MHz-50MHz BW while consuming 24μW, giving a 3-10fJ/conv-step FoM.
- 1 June 2015
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. C160-C161
A 310fJ/conv-step 0.0032mm2 error-shaping alias-free asynchronous ADC
We present a programmable, quantization error spectral shaping, alias-free asynchronous ADC suited for clockless, continuous-time DSP in receivers with modest SNDR requirements and a tight power budget. Implemented in 0.65V 28nm FDSOI, the 0.0032mm 2 ADC achieves 32dB-42dB SNDR over a 10MHz-50MHz BW while consuming 24μW, giving a 3-10fJ/conv-step FoM.
- 29 July 2014
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. 49 (10) , 2292-2304
A Flexible, Event-Driven Digital Filter With Frequency Response Independent of Input Sample Rate
This paper presents a clockless digital filter able to process inputs of different rates and formats, synchronous or asynchronous, with no adjustment needed to handle each input type. The filter is designed using a mix of asynchronous and real-time digital hardware, and for this reason relies on neither a clock nor the input data rate for setting its frequency response. The modular architecture of the filter, including delay segments with separated data and timing paths and a pipelined multi-way adder, allows easy extensions for different data widths. The filter was used as part of an ADC/DSP/DAC system which maintains its frequency response intact for varying sample rates without requiring any internal change. This property is not possible for any synchronous DSP system. The 16-tap, 8-bit FIR filter, integrated in a 130 nm CMOS process, includes on-chip automatic delay tuning, and for certain inputs, has signal-to-error ratio which exceeds that of clocked systems.
- 29 July 2014
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. 49 (10) , 2292-2304
A Flexible, Event-Driven Digital Filter With Frequency Response Independent of Input Sample Rate
This paper presents a clockless digital filter able to process inputs of different rates and formats, synchronous or asynchronous, with no adjustment needed to handle each input type. The filter is designed using a mix of asynchronous and real-time digital hardware, and for this reason relies on neither a clock nor the input data rate for setting its frequency response. The modular architecture of the filter, including delay segments with separated data and timing paths and a pipelined multi-way adder, allows easy extensions for different data widths. The filter was used as part of an ADC/DSP/DAC system which maintains its frequency response intact for varying sample rates without requiring any internal change. This property is not possible for any synchronous DSP system. The 16-tap, 8-bit FIR filter, integrated in a 130 nm CMOS process, includes on-chip automatic delay tuning, and for certain inputs, has signal-to-error ratio which exceeds that of clocked systems.
- 1 January 2008
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Solid-State Circuits Society Newsletter
- Vol. 13 (3) , 56-58
Eric Vittoz and the Strong Impact of Weak Inversion Circuits
In a groundbreaking and now classic 1977 paper, Eric Vittoz and J. Fellrath meticulously characterized and developed models for devices operating in the weak inversion region, as well as a variety of circuit building blocks that could operate and exploit the exponential characteristics in this region. Their killer application -- the electronic watch -- employed techniques that are now used in a score of low-voltage, micropower applications, such as biomedical devices, hearing aids, pagers, sensor interfaces, motion detectors for pointing devices, and a variety of portable instruments. This tribute to Vittoz as an educator and industrial researcher is by Yannis Tsividis of Columbia University, an esteemed scientist and educator in his own right.
- 1 January 2008
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Solid-State Circuits Society Newsletter
- Vol. 13 (3) , 56-58
Eric Vittoz and the Strong Impact of Weak Inversion Circuits
In a groundbreaking and now classic 1977 paper, Eric Vittoz and J. Fellrath meticulously characterized and developed models for devices operating in the weak inversion region, as well as a variety of circuit building blocks that could operate and exploit the exponential characteristics in this region. Their killer application -- the electronic watch -- employed techniques that are now used in a score of low-voltage, micropower applications, such as biomedical devices, hearing aids, pagers, sensor interfaces, motion detectors for pointing devices, and a variety of portable instruments. This tribute to Vittoz as an educator and industrial researcher is by Yannis Tsividis of Columbia University, an esteemed scientist and educator in his own right.
- 1 January 2008
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Solid-State Circuits Society Newsletter
- Vol. 13 (1) , 6-9
Turning students on to circuits
When one compares today's students to those of earlier generations, the differences are striking. Yet the way most of us teach has essentially remained unchanged since the middle of the past century. No wonder, then, that our students are not attracted to electrical engineering or, among those who are, many are disappointed and just drag along, or even drop out. In this article we discuss what can be done to turn things around. Parts of this article draw on an earlier one on an introductory EE class.
- 1 January 2008
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Solid-State Circuits Society Newsletter
- Vol. 13 (1) , 6-9
Turning students on to circuits
When one compares today's students to those of earlier generations, the differences are striking. Yet the way most of us teach has essentially remained unchanged since the middle of the past century. No wonder, then, that our students are not attracted to electrical engineering or, among those who are, many are disappointed and just drag along, or even drop out. In this article we discuss what can be done to turn things around. Parts of this article draw on an earlier one on an introductory EE class.
- 1 May 2007
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 02714302,p. 2232-2235
Analysis of Continuous-Time Digital Signal Processors
Continuous-time digital signal processors avoid sampling, and thus do not suffer from aliasing of the signal, or of aliasing down of out-of-band distortion components, arising from quantization, into the band of interest. The reduction of in-band distortion is derived analytically for certain symmetric inputs and is discussed for input processes with a Gaussian amplitude probability density function. It is demonstrated that processing in continuous time yields a 9 dB improvement of in-band signal-to-distortion ratio per added bit of quantizer resolution for both input classes examined, compared to a 6 dB per bit improvement in sampled systems.
- 1 May 2007
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 02714302,p. 2232-2235
Analysis of Continuous-Time Digital Signal Processors
Continuous-time digital signal processors avoid sampling, and thus do not suffer from aliasing of the signal, or of aliasing down of out-of-band distortion components, arising from quantization, into the band of interest. The reduction of in-band distortion is derived analytically for certain symmetric inputs and is discussed for input processes with a Gaussian amplitude probability density function. It is demonstrated that processing in continuous time yields a 9 dB improvement of in-band signal-to-distortion ratio per added bit of quantizer resolution for both input classes examined, compared to a 6 dB per bit improvement in sampled systems.