CONV-SRAM: An Energy-Efficient SRAM With In-Memory Dot-Product Computation for Low-Power Convolutional Neural Networks
Open accessThe DOI matches its Crossref record; 25 of the 29 references it cites were found in a registry or an index.
Abstract
This paper presents an energy-efficient static random access memory (SRAM) with embedded dot-product computation capability, for binary-weight convolutional neural networks. A 10T bit-cell-based SRAM array is used to store the 1-b filter weights. The array implements dot-product as a weighted average of the bitline voltages, which are proportional to the digital input values. Local integrating analog-to-digital converters compute the digital convolution outputs, corresponding to each filter. We have successfully demonstrated functionality (>98% accuracy) with the 10 000 test images in the MNIST hand-written digit recognition data set, using 6-b inputs/outputs. Compared to conventional full-digital implementations using small bitwidths, we achieve similar or better energy efficiency, by reducing data transfer, due to the highly parallel in-memory analog computations.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Among charge-based memories, SRAM cells (a bi-stable transistor structure typically made of two CMOS inverters connected back-to-back which retains a charge concentration, see Fig. 3a for an example of the structure of a crossbar array of 6T SRAM) have been widely used for VMM75–77.”
From Hardware implementation of memristor-based artificial neural networks (2024) · cites this paper as [76]
Citations
Cites
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155 · cited 1×
“Convolutional neural networks (CNNs) provide state-of-theart results in a wide variety of AI/ML applications, ranging from image classification [3] to speech recognition [1].”
BackgroundGradient-based learning applied to document recognition
Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998) · Proceedings of the IEEE · doi:10.1109/5.726791 · Cited by 58,805 · cited 5×
“However, for the MNIST dataset with LeNet-5 CNN, we estimate [22] those to have only small contributions to the overall energy-efficiency per MAV operation, due to the high parallelism supported by our in-memory approach.”
MethodInfluentialHinton, G., Deng, L., Yu, D., Dahl, G., Mohamed, A.-R., Jaitly, N., Senior, A., Vanhoucke, V., Nguyen, P., Sainath, T., et al. (2012) · IEEE Signal Processing Magazine · doi:10.1109/msp.2012.2205597 · Cited by 10,260
DeepFace: Closing the Gap to Human-Level Performance in Face Verification
Taigman, Y., Yang, M., Ranzato, M., & Wolf, L. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.220 · Cited by 6,542 · cited 1×
“Facial recognition [2] enables using our faces to get access to devices in a more intuitive manner, instead of traditional passcodes.”
BackgroundXNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks
Rastegari, M., Ordonez, V., Redmon, J., & Farhadi, A. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46493-0_32 · Cited by 4,408 · cited 2×
“[6] proposed a binary-weight-network (BWN), where the filter weights (wi s) can be trained to be +1/−1 (with a common scaling factor per 3-D filter: α).”
BackgroundMethodEfficient Processing of Deep Neural Networks: A Tutorial and Survey
Sze, V., Chen, Y.-H., Yang, T.-J., & Emer, J. S. (2017) · Proceedings of the IEEE · doi:10.1109/jproc.2017.2761740 · Cited by 4,215 · cited 1×
“In the conventional all-digital implementation of CNNs [4], [5], [9], with the memory and the processing elements being physically separate, reading wi s and the partial sums from the on-chip SRAMs lead to a lot of data movement per computation [10] and, hence, make them energy-hungry.”
BackgroundEyeriss: An Energy-Efficient Reconfigurable Accelerator for Deep Convolutional Neural Networks
Chen, Y.-H., Krishna, T., Emer, J. S., & Sze, V. (2017) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2016.2616357 · Cited by 3,247 · cited 12×
“CNNs typically consist of a cascade of convolutional (CONV) and fully-connected (FC) layers (Fig.”
BackgroundMethodInfluentialBinaryConnect: Training Deep Neural Networks with binary weights during propagations
Courbariaux, M., Bengio, Y., & David, J.-P. (2015) · arXiv · doi:10.48550/arxiv.1511.00363 · Cited by 2,180 · cited 1×
“BWN performs better than binary connect [7], which does not incorporate the scaling factor of α per filter, and also binarized neural networks [8], where both weights and activations are constrained to ±1.”
Background1.1 Computing's energy problem (and what we can do about it)
Horowitz, M. (2014) · 2014 IEEE International Solid-State Circuits Conference Digest of Technical Papers (ISSCC) · doi:10.1109/isscc.2014.6757323 · Cited by 2,069 · cited 1×
“to access data from memory can be much higher than the energy needed for a compute operation with that data [11].”
BackgroundHubara, I., Soudry, D., & Yaniv, R. E. (2016) · arXiv · doi:10.48550/arxiv.1602.02505 · Cited by 1,667 · cited 1×
“BWN performs better than binary connect [7], which does not incorporate the scaling factor of α per filter, and also binarized neural networks [8], where both weights and activations are constrained to ±1.”
Background
Research works printed but not linked: 1
- whatmough, 2017, A 28 nm SoC with a 1.2 GHz 568 nJ/prediction sparse deep-neural-network engine with >0.1 timing error rate tolerance for IoT applications, IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers, 242
Cited by
Items 301–400 of 414
Total Ionizing Dose Effects of SRAM-Based Compute-In-Memory Macro With Analog MAC Operations
Tan, J., Li, R., Ding, D., & Chen, Z. (2025) · IEEE Transactions on Nuclear Science · doi:10.1109/tns.2025.3526665 · Cited by 2
Guo, Y., Deng, L., Liu, J., Tang, C., Cheng, Y., Chen, W., Ouyang, P., Zhang, Y., Cheng, Y., & Ma, Y. (2025) · Japanese Journal of Applied Physics · doi:10.35848/1347-4065/adc8b0 · Cited by 1
SIMMAC: SRAM IMC-Based Multibit Multiplication With Analog Carry Computation
Chithambara Moorthii, J., Verma, D., Mishra, R., Bansal, H., Dey, S., Mukherjee, A., Pal, A., & Suri, M. (2025) · IEEE Embedded Systems Letters · doi:10.1109/les.2025.3559208 · Cited by 1
Quoc, D. B. L., Vo, P. L., Thien, P. N. P., & Tran, L. (2025) · Engineering Technology & Applied Science Research · doi:10.48084/etasr.11646 · Cited by 1
Liu, Y., Shang, N., Zhao, T., Chen, P., Jia, T., Wang, Z., Ma, Y., & Ye, L. (2025) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2025.3603002 · Cited by 1
“Currently, the in-memory computing (IMC) architecture prevails in terms of minimizing data movement [27], [28], [29], [30], [31].”
Khalid, A., & Ahmad, P. (2025) · Materials Today Communications · doi:10.1016/j.mtcomm.2025.113848 · Cited by 1
ROM-SRAM hybrid compute-in-memory for edge AI: circuits, architectures and challenges
Du, X., Yu, T., Zhou, H., Cheong, L.-A., Wan, T., Yang, H., & Li, X. (2025) · The Journal of Supercomputing · doi:10.1007/s11227-025-08080-2 · Cited by 2
Patel, A. A., Dasaraju, S. S., Gundrapally, A., & Choi, K. K. (2025) · Electronics · doi:10.3390/electronics14183737 · Cited by 1
Wan, R., Lin, C.-T., Zhang, B., Tsividis, Y., & Seok, M. (2025) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2025.3622309 · Cited by 1
“Recently, multiple works on CIM macro circuits have explored the potentials and challenges of analog computing [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23].”
Zhang, T., Zhu, S., Wen, Y., Pan, W., & Niu, Y. (2025) · ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp) · doi:10.1145/3714394.3756173 · Cited by 1
“However, despite the breakthroughs in hardware design of CIM chips, their application in CNN still faces many challenges, especially in convolution kernel design [9].”
Zhang, J., Tong, Z., Zhao, Q., Peng, C., Lu, W., Zhou, J., Chen, X., Lin, Z., & Wu, X. (2026) · IEEE Transactions on Very Large Scale Integration (VLSI) Systems · doi:10.1109/tvlsi.2026.3657208 · Cited by 1
“In order to further analyze the advantages of our proposed strategy in terms of linearity, we will use the strategy like [3] as a comparison.”
InfluentialChen, Y., & Xin, J. (2026) · Journal of Global Information Management · doi:10.4018/jgim.401375 · Cited by 1
Goel, A., Singhal, C., Kushwaha, D., Mittal, S., & Dasgupta, S. (2025) · IEEE Asia Pacific Conference on Circuits and Systems (APCCAS) · doi:10.1109/apccas67402.2025.11377891 · Cited by 1
A NOR8T SRAM Digital Compute-in-Memory Macro for Sparse and Scalable Edge-AI Processing
Sharma, P., Lokhande, M., Sankhe, A., Chui, K. T., Gupta, B. B., & Vishvakarma, S. K. (2025) · IEEE Asia Pacific Conference on Circuits and Systems (APCCAS) · doi:10.1109/apccas67402.2025.11377073 · Cited by 2
“Due to space constraints, we only present results for the LeNet-5 model on the MNIST dataset, demonstrating inference accuracy comparable to prior works [8], [25].”
Goel, A., Singhal, C., Mittal, S., & Dasgupta, S. (2026) · IEEE Transactions on Very Large Scale Integration (VLSI) Systems · doi:10.1109/tvlsi.2026.3663627 · Cited by 1
Zheng, L., Bavani, A. M., Chen, M., Du, S., Hsin, T.-Y., Khwa, W.-S., Wu, P.-S., Lele, A. S., Zhang, B., Crafton, B., et al. (2026) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2026.3694285 · Cited by 1
“…CMOS memories, static random access memory (SRAM)-based CIM has evolved from early in-array dot-product and vector primitives targeting CNN and DSP workloads [1], [2], [3], [4], [5], to fully-digital and multitiled implementations that push e ffi ciency frontiers in advanced nodes [6], [7], [8].”
An Area-Efficient Word-Line Pitch-Aligned 8T SRAM Compatible Digital-to-Analog Converter
Vijayakumar, B., & Viraraghavan, J. (2021) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas51556.2021.9401131 · Cited by 1
Towards Co-designing Neural Network Function Approximators with In-SRAM Computing
Nasrin, S., Badawi, D., Çetin, A. E., Gomes, W., & Trivedi, A. R. (2019) · Workshop on Energy Efficient Machine Learning and Cognitive Computing for Embedded Applications (EMC2) · doi:10.1109/emc2-nips53020.2019.00017 · Cited by 1
“, allows processing a higher number of parallel columns (l) with the same ADC complexity than in [1].”
BackgroundMethodFlashMAC: An Energy-Efficient Analog-Digital Hybrid MAC with Variable Latency-Aware Scheduling
Gweon, S., Kang, S., Han, D., Lee, K.-R., Kim, K., & Yoo, H. (2021) · IEEE Asian Solid-State Circuits Conference (A-SSCC) · doi:10.1109/a-sscc53895.2021.9634746 · Cited by 1
Wang, L., An, J., Ye, W., Li, W., Gao, H., He, Y., Gao, J., Yue, J., Fan, L., & Dou, C. (2022) · IEEE Asia Pacific Conference on Circuits and Systems (APCCAS) · doi:10.1109/apccas55924.2022.10090254 · Cited by 1
“)(2) [16] ACKNOWLEDGMENT This work was supported by the National Key R&D Program and NSFC of China under Grant No.”
BackgroundReconfigurable Compute in Memory Using SRAM and Bit Serial Algorithm
Deepa, A., Nirmalkumar, P., Jose, D., & Prasad, M. V. T. (2023) · International Conference on Inventive Research in Computing Applications (ICIRCA) · doi:10.1109/icirca57980.2023.10220833 · Cited by 2
“It also consist of the MAC unit output of P[0], P[1], P[2], P[3], P[4].”
BackgroundMethodInfluentialVenu, B., Rao, K. T., Penumalli, K., Japa, A., Sambatur, S. N., Gu, C., Yellampalli, S. S., & Vaddi, R. (2024) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas58744.2024.10558683 · Cited by 2
“The authors in [21] have been presented the convolution (conv) SRAM for a machine learning-based dot product system.”
MethodWu, H., Lin, Z., Wu, X., Zhao, Q., Lu, W., & Peng, C. (2024) · International Journal of Circuit Theory and Applications · doi:10.1002/cta.4192 · Cited by 1
BP-IMCA: An Energy-Efficient 8T SRAM-Based Bit-Parallel In-Memory Computing Architecture
Rajput, A. K., Pattanaik, M., & Kaushal, G. (2024) · Journal of Circuits Systems and Computers · doi:10.1142/s0218126625501245 · Cited by 1
An Energy Efficient C-2C Charge-Sharing Based Analog Compute-In-Memory Architecture
Kushwaha, D., Kohli, R., Mishra, J., Singh, J., Joshi, R., Dasgupta, S., & Bulusu, A. (2023) · IEEE Asia Pacific Conference on Postgraduate Research in Microelectronics and Electronics (PrimeAsia) · doi:10.1109/primeasia60757.2023.00017 · Cited by 1
“The concept of compute-in-memory (CIM) architecture is evolved in recent years to solve the limitation of conventional digital computing architectures [6-25].”
BackgroundGundrapally, A., Alnatsheh, N., & Choi, K. K. (2024) · Electronics · doi:10.3390/electronics13234828 · Cited by 1
ADDR: Architecture Design and Model Deployment Optimization for Hybrid SRAM-ROM Compute-in-Memory
Wan, T., Chen, Y., Chen, Z., Liu, Y., Yang, H., & Li, X. (2025) · ACM Great Lakes Symposium on VLSI (GLSVLSI) · doi:10.1145/3716368.3735252 · Cited by 1
Selvarasu, S., & Saravanan, S. (2019) · Soft Computing · doi:10.1007/s00500-019-04581-4 · Cited by 1
RetractedAn In-Memory Architecture for Machine Learning Classifier Using Logistic Regression
Saragada, P. K., Rathod, M., & Das, B. P. (2019) · IEEE International Symposium on Smart Electronic Systems (iSES) · doi:10.1109/ises47678.2019.00054 · Cited by 1
“However, on-chip training of machine learning classifier is not implemented in these work [6-7].”
BackgroundMethodInfluential10.76 TOPS/W CNN Algorithm Circuit using Processor-In-Memory with 8T-SRAM
Park, K. D., Kim, D., Pu, Y., & Lee, K. (2021) · IEEE International Conference on Big Data and Smart Computing (BigComp) · doi:10.1109/bigcomp51126.2021.00071 · Cited by 1
An Automated Design Methodology for Computational SRAM Dedicated to Highly Data-Centric Applications
Philippe, A., Ciampolini, L., Philippe, A., Gerbaud, M., Ramirez-Corrales, M., Egloff, V., Giraud, B., & Noël, J.-P. (2022) · doi:10.1145/3557988.3569715 · Cited by 1
“This is commonly referred to as the in-and near-memory computing (IMC/NMC) paradigm [4]-[6].”
Accelerating Low Bit-width Neural Networks at the Edge, PIM or FPGA: A Comparative Study
Kochar, N., Ekiert, L., Najafi, D., Fan, D., & Angizi, S. (2023) · ACM Great Lakes Symposium on VLSI (GLSVLSI) · doi:10.1145/3583781.3590213 · Cited by 2
Edge processing in IoT using approximate and in-memory computing
Bose, S. K. (2021) · DR-NTU (Nanyang Technological University) · doi:10.32657/10356/152016 · Cited by 1
Hardware and algorithm co-optimization for energy-efficient machine learning integrated circuits
Kim, H. (2022) · DR-NTU (Nanyang Technological University) · doi:10.32657/10356/168401 · Cited by 1
Energy-Efficient Hybrid-RAM with Hybrid Bit-Serial based VMM Support
Nie, C., Lin, J., Hu, H., Jiang, L., Liang, X., & He, Z. (2021) · ACM Great Lakes Symposium on VLSI (GLSVLSI) · doi:10.1145/3453688.3461528
Rajput, A. K., Pattanaik, M., & Kaushal, G. (2021) · doi:10.36227/techrxiv.16704805.v1
Temporal Accelerators: Unleashing the Potential of Embedded FPGAs
Cichiwskyj, C., & Schiele, G. (2021) · JUCS - Journal of Universal Computer Science · doi:10.3897/jucs.77247
Rajput, A. K., Pattanaik, M., & Kaushal, G. (2021) · doi:10.36227/techrxiv.16704805
“In this section, the benchmarks circuits of C6T, Transpose-8T [15], DP8T [16], 8T [11] 10T [20] and the proposed circuit are implemented with the same 65nm CMOS technology with exact sizing and simulate at 1V supply voltage and 27 ◦ C temperature with a typical process corner for performance…”
BackgroundMethodResultInfluentialMemory and Energy Efficient Memory Model and Instruction Set Architectures for Tree Data Structures
Rifai, M., & Johnsson, L. (2022) · International Symposium on Highly Efficient Accelerators and Reconfigurable Technologies (HEART) · doi:10.1145/3535044.3535052
“Both near-memory and bit-serial computing are used with software algorithms ranging from neural networks [7, 10, 16, 20, 40] to graph and signal processing.”
MethodYu, C., Kim, H., Kim, B., & Kim, T. T.-H. (2022) · Springer eBooks · doi:10.1007/978-3-030-98781-7_2
Input Conditioned Subranging and Skewed Quantisation of MACs in IMCPreprint
Sundar, A. B., Viraraghavan, J., & Vijayakumar, B. (2022) · doi:10.36227/techrxiv.20330580.v1
Input Conditioned Subranging and Skewed Quantisation of MACs in IMCPreprint
Sundar, A. B., Viraraghavan, J., & Vijayakumar, B. (2022) · doi:10.36227/techrxiv.20330580
A DCSRO Based Time Domain MAC CorePreprint
Musah, T., Du, K., & Abdelaziz, A. (2023) · doi:10.22541/au.167710265.55273622/v1
A 0.32V Bit-Interleave 8T SRAM based In-Memory Computing Macro for AI Edge DevicesPreprint
Rajput, A. K., Pattanaik, M., Kaushal, G., & Vishvakarma, S. K. (2023) · doi:10.36227/techrxiv.22739633
A 0Preprint
Rajput, A. K., Pattanaik, M., Kaushal, G., & Vishvakarma, S. K. (2023) · doi:10.36227/techrxiv.22739633.v1
Du, K., Abdelaziz, A., & Musah, T. (2023) · International Journal of Circuit Theory and Applications · doi:10.1002/cta.3717
Acacia Senegal-based Non-Newtonian Electrolyte for Memristive ApplicationsPreprint
Lokhandwala, A., Thakkar, P., Gosai, J., Oza, S., & Solanki, A. (2023) · Research Square · doi:10.21203/rs.3.rs-3105809/v1
SRAM-Based In-Memory Computing: Circuits, Functions, and Applications
Hassan, E., Tesfai, H., Mohammad, B. S., & Saleh, H. H. (2012) · Springer eBooks · doi:10.1007/978-3-031-34233-2_2 · Cited by 1
Robustness for Embedded Machine Learning Using In-Memory Computing
Panda, P., Bhattacharjee, A., & Moitra, A. (2023) · Springer eBooks · doi:10.1007/978-3-031-40677-5_17
Hong, H.-C., Chen, C.-H., & Chen, Y.-W. (2023) · IEEE International Test Conference in Asia (ITC-Asia) · doi:10.1109/itc-asia58802.2023.10301157
“1: Conceptual CIM scheme. the memory cells [2]–[6].”
BackgroundDecoupling efficiency-performance optimization for modern neural networks
Jin, Q. (2023) · Digital Repository Service (Northeastern University) · doi:10.17760/d20592335
In‐memory multibit multiplication and accumulation based on an automatic pulse generation circuit
Bai, S., Guo, J., Liu, Y., Xu, Y., Peng, C., & Wu, X. (2023) · Electronics Letters · doi:10.1049/ell2.13036
Chen, Y., Yin, G., Zhong, H., Lee, M., Yang, H., George, S., Narayanan, V., & Li, X. (2024) · Asia and South Pacific Design Automation Conference (ASP-DAC) · doi:10.1109/asp-dac58780.2024.10473851
“CIM has been explored in various memory techniques, including SRAM [3], [4], eDRAM [5], [6], RRAM [7], FeFET [8], and even ROM [9].”
MethodZhu, Y., & Ohsawa, T. (2024) · IEICE Transactions on Electronics · doi:10.1587/transele.2023ecp5051
“In Table 3, bit cell area, throughput, area efficiency and energy efficiency in the papers [11 , 28, 29] were scaled from 65nm with individual V dd to 32nm with V dd =1.2V according to the scaling equations proposed in paper [30].”
MethodTNN-CIM: An In-SRAM CMOS Implementation of TNN-Based Synaptic Arrays with STDP Learning
Nair, H., Barajas-Jasso, D., Jacobson, Q. A., & Shen, J. P. (2024) · IEEE International Conference on Artificial Intelligence Circuits and Systems (AICAS) · doi:10.1109/aicas59952.2024.10595925
Variation-Aware Design Methodology for SRAM-Based Multi-Bit Analog Compute-in-Memory Architecture
Kushwaha, D., Joshi, R., Bulusu, A., & Dasgupta, S. (2024) · IEEE Interregional NEWCAS Conference (NEWCAS) · doi:10.1109/newcas58973.2024.10666330
Li, F., Xia, Y., Xiao, S., Yang, P., Zhu, X., Li, B., Zhou, J., Han, G., & Mao, W. (2024) · IEEE International System-on-Chip Conference (SOCC) · doi:10.1109/socc62300.2024.10737831
“To address this issue, the computing-in-memory (CIM) architecture integrates memory and compute units for data transport reduction [3]–[5].”
MethodAndo, S., Chen, Y.-C., Miyagi, S., Zhang, W., & Yoshioka, K. (2024) · Japanese Journal of Applied Physics · doi:10.35848/1347-4065/ad9611
Dou, S., You, H., Zhan, Y., Qiao, S., & Zhou, Y. (2024) · IEICE Electronics Express · doi:10.1587/elex.21.20240694
Han, G., Chen, Z., Chen, Y., Chen, Z., Wan, P., & Liu, X. (2024) · IEEE International Conference on Anti-counterfeiting, Security, and Identification (ASID) · doi:10.1109/asid63618.2024.10839666
High-Endurance Contention-Aware Power Optimized Hybrid Memory Using Clock Gating Memory
Narayana Murty, N., Singh, H., & Thethi, S. K. (2024) · IEEE International Conference of Electron Devices Society Kolkata Chapter (EDKCON) · doi:10.1109/edkcon62339.2024.10870768
“There are many items that use delay buffers—also called line buffers or delay lines—in their circuitry [1]–[3].”
A Computing-In-Memory Analog Popcount Circuit Utilizing 9T-SRAM
Park, C., Yoo, M., & Burm, J. (2025) · International Conference on Electronics, Information, and Communication (ICEIC) · doi:10.1109/iceic64972.2025.10879686
Su, X., & Mai, Y. (2025) · Lecture notes in electrical engineering · doi:10.1007/978-981-96-5318-8_9
MOSAIC: Collaborative Compute-in-Memory µArrays for Flexible and Scalable Deep Learning
Trivedi, A. R., Nasrin, S., Shukla, P., Darabi, N., Kumar, D., Jayasuriya, D., & Ns, J. (2025) · IEEE VLSI Test Symposium (VTS) · doi:10.1109/vts65138.2025.11022889
“This enables computations on a low-capacitance pipeline, unlike conventional large-scale processing with higher overhead [8], [9].”
In-Memory Logic Synthesis Methods based on Read Decoupled 8T and Transpose 11T SRAMs
Wei, F., Cui, X., & Zhang, S. (2025) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas56072.2025.11043683
Alorf, A., Taylor, B., & Chen, Y. (2025) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas56072.2025.11043903
“Operating in the analog domain [15], [17]–[20] has been shown to be a viable solution, offering more efficient designs, especially in terms of power and area, compared to the digital domain.”
Liu, Z., Zhang, Y., Zhang, Z., Yang, Y. T., Zhang, Z., Wang, X., Chen, J., Li, X., Si, X., & Yang, J. (2025) · Science China Information Sciences · doi:10.1007/s11432-024-4354-y
A Reconfigurable Variable-Precision Bipolar/Binary QR-based IMC Accelerator in 22nm FDSOI
Quizon, L. R. A., Baquiran, S. J. A. S., & Alvarez, A. B. (2025) · International Conference on Advanced Machine Learning and Data Science (AMLDS) · doi:10.1109/amlds63918.2025.11159452
“The first significant QR-based IMC was presented in 2018 with Conv-RAM [5].”
Lee, S., Yi, D., Choi, E. J., Lukito, V., Kim, J.-H., Koh, R. G. L., Ha, S., Chang, I., & Je, M. (2025) · IEEE International Conference on Artificial Intelligence Circuits and Systems (AICAS) · doi:10.1109/aicas64808.2025.11173134
“The design exploits the integration of an analog compute-in-memory (CiM) [13]–[21] technique based on magnetoresis-tive random-access memory (MRAM) [22]–[24] to implement a customized DNN, eliminating the need for external processors, as shown in Fig.”
Dual-path synchronized discharge CIM architecture with bipolar computing unit for BWNs
Zhou, Y., Tan, W., Yang, Z., Peng, C., & Wu, X. (2025) · Microelectronics Journal · doi:10.1016/j.mejo.2025.106899
A Booth‐Encoded Digital Computing‐in‐Memory Macro With Configurable Approximate Compressor Tree
Pan, B., Ren, Y., Gao, Z., Wei, T., Wu, H. S., & He, Y. (2025) · Electronics Letters · doi:10.1049/ell2.70434
ThRIve: Thermally Robust CNN Inference via Low-Rank Adaptation in Heterogeneous PIM Architectures
Sharma, V., Dhingra, P., Doppa, J. R., & Pande, P. P. (2025) · ACM Transactions on Design Automation of Electronic Systems · doi:10.1145/3774328 · Cited by 1
“PIM implementations employ both CMOS-based volatile memory such as Static Random-Access Memory (SRAM) and Non-Volatile Memory (NVM) devices such as Resistive Random-Access Memory (ReRAM), Phase Change Memory (PCM), and Ferroelectric Field-Effect Transistors (FeFET) etc. [7], [8], [9], [10], [11].”
Automatic Verification of Analog and Mixed-Signal Neural Network Accelerators and Matrix Multipliers
Müller, R., Kundu, B., Vieregge, J. M., Patil, Y., & Brederlow, R. (2025) · IEEE Nordic Circuits and Systems Conference (NorCAS) · doi:10.1109/norcas66540.2025.11231195
Analyzing Embedded Non-Volatile Memory and Technology Optimizations for In-Memory Computing
Deaville, P., Zhang, B., & Verma, N. (2025) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2025.3632212
“V conversion [12], [30], and they all invoke the same fundamental principle of establishing an IMC signal above the noise of the readout circuitry.”
A CIM Crossbar Array Data Mapping Methodology for Unstructured Sparse CNN
Hung, Y.-L., Liu, P.-H., Wu, Z.-S., Lai, B.-C., & Jou, S.-J. (2025) · IEEE International Conference on Artificial Intelligence Circuits and Systems (AICAS) · doi:10.1109/aicas64808.2025.11173116
“Compute-In-Memory (CIM) [10] has been introduced alleviate this issue by performing computing on stored data within a crossbar memory array structure, to not only reduce the data transactions but also enables massively parallel processing with low energy consumption [11, 12].”
A Robust 4T1C eDRAM Compute-in-Memory Architecture for Inference Applications
Singh, A. K., Kushwaha, D., & Dasgupta, S. (2025) · IEEE Interregional NEWCAS Conference (NEWCAS) · doi:10.1109/newcas64648.2025.11106976
“The input activation (x) is converted into analog input voltage (V a ) through the digital-to-analog converter (DAC) [20] and pre-charge the read bitline (RBL).”
Kushwaha, D., Dasgupta, S., & Bulusu, A. (2025) · IEEE Interregional NEWCAS Conference (NEWCAS) · doi:10.1109/newcas64648.2025.11106996
“…(cid:24)(cid:24)(cid:25) ∑ (cid:26) (cid:24)(cid:31)(cid:30) (4) In this work the digital to analog converter (DAC ) is borrowed from [4] whereas the readout circuit consist of 4-bit flash ADC, as reported in our previous work [16].”
InfluentialNovel Low-Power CNFET-GAAFET Based Ternary 9T SRAM Design for Computing-in-Memory Systems
Patel, A. A., Dasaraju, S. S., Shah, Y. A., Gundrapally, A., & Choi, K. K. (2025) · Electronics · doi:10.3390/electronics15010137
Lin, Z., Duan, C., Long, M., Ge, J., Jiang, F., Yang, Y., & Wu, X. (2025) · International Conference on Integrated Circuits and Microsystems (ICICM) · doi:10.1109/icicm66614.2025.11315968
“…memory arrays, CIM greatly reduces data movement and thereby offers significant gains in both energy efficiency and processing speed. for key arithmetic tasks such as multiply-accumulate (MAC) operations [5], [6], efficiently implementing Boolean logic operations in memory remains a challenge.”
A Stochastic Computing-Based Computing-in-Memory Macro with Bit-Split Stochastic Number Generation
Lin, Z., Chen, D., Li, S., Liu, S., Liu, Y., & Wu, X. (2025) · International Conference on Integrated Circuits and Microsystems (ICICM) · doi:10.1109/icicm66614.2025.11315958
“To address the limitations inherent in traditional von Neumann architecture, particularly the memory wall issue, the paradigm of Computing-In-Memory (CIM) [1]–[3] has been proposed.”
An Efficient-Energy Charge-Domain Convolution Operator for CNN
Diaz-Madrid, J.-A., Doménech‐Asensi, G., Ruíz, R., & Andreo-Oliver, F. J. (2026) · IEEE Access · doi:10.1109/access.2026.3657997
“In [5], the incoming digital signal is converted into an analog signal, serving as a precharge for the bit lines.”
Hardware-efficient custom floating-point dot-product processor for LSTM
Venkatesh, S., Sindhu, R., Hallur, A. G., Sharma, K. S., & Arunachalam, V. (2026) · International Journal of Electronics Letters · doi:10.1080/21681724.2026.2620382
“LSTM networks can be deployed efficiently thanks to optimised floating-point designs, fused dot-product units, and SRAM-based in-memory computation, which further reduce power, latency, and area (Biswas & Chandrakasan, 2018; Hickmann et al., 2020; Li et al., 2023).”
Jiang, L., Wu, J., Sun, S., Li, C., Kang, W., & Zhang, H. (2026) · IEEE Transactions on Very Large Scale Integration (VLSI) Systems · doi:10.1109/tvlsi.2026.3656426
“During the bu ff er phase, these analog voltages are then held in the bu ff er until all signals are synchronized [16], [18], [20], [21].”
Min, J. H., Kang, J., Kim, J. H., & Cho, S. (2026) · IEEE Access · doi:10.1109/access.2026.3660349
“The first approach is the current-domain method [16], [17], [18], in which the multiplication result between the input activation and the corresponding weight determines the discharge current, and these results are simultaneously accumulated on the bit line (BL).”
Test System for an SRAM-Based Mixed-Signal Accelerator
Caselli, M., Malena, F., & Boni, A. (2026) · Lecture notes in electrical engineering · doi:10.1007/978-3-032-17174-0_23
Li, X., Zhang, W., Ando, S., Wang, Z., Lin, J., & Yoshioka, K. (2026) · IEEE Transactions on Very Large Scale Integration (VLSI) Systems · doi:10.1109/tvlsi.2026.3665471
“In [31], a 6b input × 1b weight charge-sharing architecture is first proposed.”
MemSOS: OS-Guided Selective Memory Mirroring
Kim, J., Jeong, J., Moon, S., Seo, S. H., Park, Y., Jeong, J., Kim, N., & Lee, J. W. (2026) · IEEE International Symposium on High-Performance Computer Architecture (HPCA) · doi:10.1109/hpca68181.2026.11408560
“Using CACTI-P [47], we evaluate their overhead at 40nm and scale to 7nm following Biswas and Chan-drakasan [3].”
Yao, J., Yang, Z., Jiang, Y., Yang, Y., Yan, Z., Zhuo, C., & Yin, X. (2026) · Asia and South Pacific Design Automation Conference (ASP-DAC) · doi:10.1109/asp-dac66049.2026.11420362
Readout scheme in analog computing-in-memory: Review of design principles and recent progress
Zhang, S., Huang, P., You, H., Wu, X., Chen, W., Zhang, L., & Cui, Q. (2026) · Microelectronic Engineering · doi:10.1016/j.mee.2026.112466
Number Plate Recognition Using Machine Learning Algorithms and CNN
Pandya, V. C., Savani, V., Amrutiya, Y., Adatiya, Y., & Padariya, D. (2026) · Smart innovation, systems and technologies · doi:10.1007/978-3-032-07837-7_1
In-memory computing based on CMOS-RRAM hybrid architectures
Esmanhotto, E. (2022) · theses.fr (ABES) · doi:10.70675/1cf79608z3789z4025z8efcz817e8621bdd9 · Cited by 1
Gautier, A. (2024) · theses.fr (ABES) · doi:10.70675/5827871fz5213z492ez9180z641a7bd710e5
CASTRA: A Digital Processing-In-situ SRAM Macro for Resource-Efficient TinyML AccelerationPreprint
Lokhande, M., Sankhe, A., Teman, A., & Vishvakarma, S. K. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6663807
Singhal, C., Goel, A., Shahi, K., Palak, Mittal, S., & Dasgupta, S. (2026) · International Symposium on VLSI Technology, Systems and Applications (VLSI-TSA) · doi:10.1109/vlsitsa69131.2026.11527555
Kumar, C. P., Santosh, B. C. R. S., Kumar, C. S. M., Mathan, N., & Reddy, C. A. M. (2026) · International Conference on Communications and Signal Processing (ICCSP) · doi:10.1109/iccsp68173.2026.11539240
Vishwakarma, V., Raut, G., Mittal, A., Goyal, N., Mohammad, B. S., & Vishvakarma, S. K. (2026) · IEEE Transactions on Very Large Scale Integration (VLSI) Systems · doi:10.1109/tvlsi.2026.3697581
“The prior works include recent digital and mixed-signal SRAM-CIM implementations reported in TCAS-I’24 [12], TCAS’25 [15], JSSC’19 [33], JETCAS’22 [34], TNANO’23 [35], JSSC’21 [19], and JSSC’24 [22].”
InfluentialLow-Complexity Rule-Based Greedy Optimization of Refractory Periods for In-Memory SNN
Tushar, S. N. B., Alam, S. H., Chakraborty, N. N., Schuman, C. D., Das, H., & Rose, G. S. (2026) · IEEE Dallas Circuits and Systems Conference (DCAS) · doi:10.1109/dcas69364.2026.11544440
“However, the energy cost associated with data movement between memory and processing units–the von Neumann bottleneck–has motivated the adoption of in-memory computing architectures [4], [5], [6].”
Investigating Energy Bounds of Analog Compute-in-Memory With Local Normalization
Rojkov, B., Ranjan, S., Wright, D., & Sachdev, M. (2026) · IEEE Journal on Emerging and Selected Topics in Circuits and Systems · doi:10.1109/jetcas.2026.3702517
Metadata
- Authors
- Biswas, A., & Chandrakasan, A. P.
- Affiliations
- Texas Instruments (United States) (US)
- Massachusetts Institute of Technology (US)
- Venue
- IEEE Journal of Solid-State Circuits
- Published
- 2019-01
- Volume
- 54
- Issue
- 1
- Pages
- 217-230
- ISSN
- 0018-9200 (ISSN-L)
- 1558-173X
- OpenAlex
- W2904299207
- Cited by
- 414
- References
- 29
- Citations by year
- 2026: 36 · 2025: 57 · 2024: 75 · 2023: 75 · 2022: 74 · 2021: 59 · 2020: 24 · 2019: 13
- FWCI
- 9.17
- Type
- journal article
- Language
- en
- Open access
- green
- hdl.handle.net
- Licence
- Locations
- IEEE Journal of Solid-State Circuits (publishedVersion)
- DSpace@MIT (Massachusetts Institute of Technology) (submittedVersion, cc-by-nc-sa)
- Funders
- Intel Corporation
- Topics
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Advanced Memory and Neural Computing
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Low-power high-performance VLSI design
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Analog and Mixed-Signal Circuit Design
- Keywords
- SRAM, in-memory computing, dot product computation, energy efficiency, low-power neural network hardware, analog-to-digital converter, MNIST classification
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.