New Modified 4:2 Approximate Compressors for Low-power Applications

被引:1
|
作者
Zanandrea, Vinicius [1 ]
Meinhardt, Cristina [1 ]
机构
[1] Univ Fed Santa Catarina, Dept Informat & Estat, PPGCC, Florianopolis, Brazil
关键词
approximate computing; energy efficiency; compressors; multipliers; DESIGN; MULTIPLIERS;
D O I
10.1109/SBCCI60457.2023.10261949
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
Energy efficiency is an important design concern in most digital and embedded systems. In particular, hardware projects often require high-speed and low-power multipliers. Compressors usually implement the partial products reduction of a multiplier to reduce the critical path and the power consumption. An emerging approach that can be applied in error-tolerant applications to improve the overall performance is the approximate computing. In this paper, two power-efficient 4:2 approximate compressors are proposed and evaluated at the transistor level in a 7 nm FinFET technology. The proposed circuits show improvements in power and delay compared to state-of-the-art approximate compressors, reducing the PDP on average by 66.7% and 79.1%, respectively. Furthermore, operating at near-threshold voltage, the energy efficiency can be improved by 43% and 52% compared to nominal operation.
引用
收藏
页码:209 / 214
页数:6
相关论文
共 50 条
  • [31] New smart power modules for very low-power drive applications
    Lee, JB
    Jang, KY
    Chung, DW
    Suh, BS
    PESC'03: 2003 IEEE 34TH ANNUAL POWER ELECTRONICS SPECIALISTS CONFERENCE, VOLS 1-4, CONFERENCE PROCEEDINGS, 2003, : 436 - 441
  • [32] Novel Architecture for Low-Power CNTFET-Based Compressors
    Gavaber, Morteza Dadashi
    Poorhosseini, Mehrdad
    Mozafari, Saadat Pour
    JOURNAL OF CIRCUITS SYSTEMS AND COMPUTERS, 2019, 28 (12)
  • [33] Low-power and variation-aware approximate arithmetic units for Image Processing Applications
    Mirzaei, Mohammad
    Mohammadi, Siamak
    AEU-INTERNATIONAL JOURNAL OF ELECTRONICS AND COMMUNICATIONS, 2021, 138
  • [34] ATLAS: An Approximate Time-Series LSTM Accelerator for Low-Power IoT Applications
    Kress, Fabian
    Serdyuk, Alexey
    Hiegle, Micha
    Waldmann, Disnebio
    Hotfilter, Tim
    Hoefer, Julian
    Hamann, Tim
    Barth, Jens
    Kaempf, Peter
    Harbaum, Tanja
    Becker, Juergen
    2023 26TH EUROMICRO CONFERENCE ON DIGITAL SYSTEM DESIGN, DSD 2023, 2023, : 569 - 576
  • [35] Machine Learning based Power Efficient Approximate 4: 2 Compressors for Imprecise Multipliers
    Maddisetti, Lavanya
    Ravindra, J. V. R.
    2019 32ND INTERNATIONAL CONFERENCE ON VLSI DESIGN AND 2019 18TH INTERNATIONAL CONFERENCE ON EMBEDDED SYSTEMS (VLSID), 2019, : 221 - 226
  • [36] Novel Architectures for high-speed and low-power 3-2, 4-2 and 5-2 compressors
    Veeramachaneni, Sreehari
    Kirthi Krishna, M.
    Avinash, Lingamneni
    Puppala, Sreekanth Reddy
    Srinivas, M. B.
    20TH INTERNATIONAL CONFERENCE ON VLSI DESIGN, PROCEEDINGS: TECHNOLOGY CHALLENGES IN THE NANOELECTRONICS ERA, 2007, : 324 - +
  • [37] Low-Power Modified Vedic Multiplier
    Kodali, Ravi Kishore
    Sivakumar, C.
    Jain, Vishal
    Boppana, Lakshmi
    2015 INTERNATIONAL CONFERENCE ON CONTROL COMMUNICATION & COMPUTING INDIA (ICCC), 2015, : 454 - 458
  • [38] Low-Power Ternary Multiplication Using Approximate Computing
    Nagesh, Gode
    Sumalatha, Madipalli
    2024 INTERNATIONAL CONFERENCE ON SOCIAL AND SUSTAINABLE INNOVATIONS IN TECHNOLOGY AND ENGINEERING, SASI-ITE 2024, 2024, : 208 - 212
  • [39] A Low-Power Approximate Multiply-Add Unit
    Yang, Tongxin
    Sato, Toshinori
    Ukezono, Tomoaki
    2019 2ND INTERNATIONAL SYMPOSIUM ON DEVICES, CIRCUITS AND SYSTEMS (ISDCS 2019), 2019,
  • [40] Low-power digital filtering using approximate processing
    Ludwig, JT
    Nawab, SH
    Chandrakasan, AP
    IEEE JOURNAL OF SOLID-STATE CIRCUITS, 1996, 31 (03) : 395 - 400