E-MAC: Enhanced In-SRAM MAC Accuracy via Digital-to-Time Modulation
In this article, we introduce a novel technique called E-multiplication and accumulation (MAC) (EMAC), aimed at enhancing energy efficiency, reducing latency, and improving the accuracy of analog-based in-static random access memory (SRAM) MAC accelerators. Our approach involves a digital-to-time wo...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Journal on Exploratory Solid-State Computational Devices and Circuits |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10804123/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832590329657688064 |
|---|---|
| author | Saeed Seyedfaraji Salar Shakibhamedan Amire Seyedfaraji Baset Mesgari Nima Taherinejad Axel Jantsch Semeen Rehman |
| author_facet | Saeed Seyedfaraji Salar Shakibhamedan Amire Seyedfaraji Baset Mesgari Nima Taherinejad Axel Jantsch Semeen Rehman |
| author_sort | Saeed Seyedfaraji |
| collection | DOAJ |
| description | In this article, we introduce a novel technique called E-multiplication and accumulation (MAC) (EMAC), aimed at enhancing energy efficiency, reducing latency, and improving the accuracy of analog-based in-static random access memory (SRAM) MAC accelerators. Our approach involves a digital-to-time word-line (WL) modulation technique that encodes the WL voltage while preserving the necessary linear voltage drop for precise computations. This eliminates the need for an additional digital-to-analog converter (DAC) in the design. Furthermore, the SRAM-based logical weight encoding scheme we present reduces the reliance on capacitance-based techniques, which typically introduce area overhead in the circuit. This approach ensures consistent voltage drops for all equivalent cases [i.e., <inline-formula> <tex-math notation="LaTeX">$(a { \times} b) = (b \times a)$ </tex-math></inline-formula>], addressing a persistent issue in existing state-of-the-art methods. Compared with state-of-the-art analog-based in-SRAM techniques, our E-MAC approach demonstrates significant energy savings (<inline-formula> <tex-math notation="LaTeX">$1.89\times $ </tex-math></inline-formula>) and improved accuracy (73.25%) per MAC computation from a 1-V power supply, while achieving a <inline-formula> <tex-math notation="LaTeX">$11.84\times $ </tex-math></inline-formula> energy efficiency improvement over baseline digital approaches. Our application analysis shows a marginal overall reduction in accuracy, i.e., a 0.1% and 0.17% reduction for LeNet5-based CNN and VGG16, respectively, when trained on the MNIST and ImageNet datasets. |
| format | Article |
| id | doaj-art-628a643239e948d5987f214ebc6a979f |
| institution | Kabale University |
| issn | 2329-9231 |
| language | English |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal on Exploratory Solid-State Computational Devices and Circuits |
| spelling | doaj-art-628a643239e948d5987f214ebc6a979f2025-01-24T00:02:11ZengIEEEIEEE Journal on Exploratory Solid-State Computational Devices and Circuits2329-92312024-01-011017818610.1109/JXCDC.2024.351863310804123E-MAC: Enhanced In-SRAM MAC Accuracy via Digital-to-Time ModulationSaeed Seyedfaraji0https://orcid.org/0000-0003-0085-6282Salar Shakibhamedan1https://orcid.org/0000-0003-2862-2859Amire Seyedfaraji2https://orcid.org/0000-0002-7047-9454Baset Mesgari3https://orcid.org/0000-0002-2109-8348Nima Taherinejad4https://orcid.org/0000-0002-1295-0332Axel Jantsch5https://orcid.org/0000-0003-2251-0004Semeen Rehman6Technische Universität Wien (TU Wien), Vienna, AustriaTechnische Universität Wien (TU Wien), Vienna, AustriaDepartment of Electrical Engineering, Faculty of Engineering, Alzahra University, Tehran, IranTechnische Universität Wien (TU Wien), Vienna, AustriaTechnische Universität Wien (TU Wien), Vienna, AustriaTechnische Universität Wien (TU Wien), Vienna, AustriaInstitute of Parallel Computing Systems, Computer Science, University of Amsterdam, Amsterdam, The NetherlandsIn this article, we introduce a novel technique called E-multiplication and accumulation (MAC) (EMAC), aimed at enhancing energy efficiency, reducing latency, and improving the accuracy of analog-based in-static random access memory (SRAM) MAC accelerators. Our approach involves a digital-to-time word-line (WL) modulation technique that encodes the WL voltage while preserving the necessary linear voltage drop for precise computations. This eliminates the need for an additional digital-to-analog converter (DAC) in the design. Furthermore, the SRAM-based logical weight encoding scheme we present reduces the reliance on capacitance-based techniques, which typically introduce area overhead in the circuit. This approach ensures consistent voltage drops for all equivalent cases [i.e., <inline-formula> <tex-math notation="LaTeX">$(a { \times} b) = (b \times a)$ </tex-math></inline-formula>], addressing a persistent issue in existing state-of-the-art methods. Compared with state-of-the-art analog-based in-SRAM techniques, our E-MAC approach demonstrates significant energy savings (<inline-formula> <tex-math notation="LaTeX">$1.89\times $ </tex-math></inline-formula>) and improved accuracy (73.25%) per MAC computation from a 1-V power supply, while achieving a <inline-formula> <tex-math notation="LaTeX">$11.84\times $ </tex-math></inline-formula> energy efficiency improvement over baseline digital approaches. Our application analysis shows a marginal overall reduction in accuracy, i.e., a 0.1% and 0.17% reduction for LeNet5-based CNN and VGG16, respectively, when trained on the MNIST and ImageNet datasets.https://ieeexplore.ieee.org/document/10804123/6T-static random access memory (SRAM)convolutional neural network (CNN)image classificationprocessing in memory (PIM) |
| spellingShingle | Saeed Seyedfaraji Salar Shakibhamedan Amire Seyedfaraji Baset Mesgari Nima Taherinejad Axel Jantsch Semeen Rehman E-MAC: Enhanced In-SRAM MAC Accuracy via Digital-to-Time Modulation IEEE Journal on Exploratory Solid-State Computational Devices and Circuits 6T-static random access memory (SRAM) convolutional neural network (CNN) image classification processing in memory (PIM) |
| title | E-MAC: Enhanced In-SRAM MAC Accuracy via Digital-to-Time Modulation |
| title_full | E-MAC: Enhanced In-SRAM MAC Accuracy via Digital-to-Time Modulation |
| title_fullStr | E-MAC: Enhanced In-SRAM MAC Accuracy via Digital-to-Time Modulation |
| title_full_unstemmed | E-MAC: Enhanced In-SRAM MAC Accuracy via Digital-to-Time Modulation |
| title_short | E-MAC: Enhanced In-SRAM MAC Accuracy via Digital-to-Time Modulation |
| title_sort | e mac enhanced in sram mac accuracy via digital to time modulation |
| topic | 6T-static random access memory (SRAM) convolutional neural network (CNN) image classification processing in memory (PIM) |
| url | https://ieeexplore.ieee.org/document/10804123/ |
| work_keys_str_mv | AT saeedseyedfaraji emacenhancedinsrammacaccuracyviadigitaltotimemodulation AT salarshakibhamedan emacenhancedinsrammacaccuracyviadigitaltotimemodulation AT amireseyedfaraji emacenhancedinsrammacaccuracyviadigitaltotimemodulation AT basetmesgari emacenhancedinsrammacaccuracyviadigitaltotimemodulation AT nimataherinejad emacenhancedinsrammacaccuracyviadigitaltotimemodulation AT axeljantsch emacenhancedinsrammacaccuracyviadigitaltotimemodulation AT semeenrehman emacenhancedinsrammacaccuracyviadigitaltotimemodulation |