HiT-CIM: A High-Throughput Compute-in-Memory SRAM Architecture With Simultaneous Weight Loading/Computing and Balance Capabilities
- 9 October 2024
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Emerging Topics in Computing
- Vol. 13 (4) , 1396-1409
Abstract
In the post-Moore's era, compute-in-memory (CIM) techniques are promising to break the memory wall. In particular, SRAM-based CIMs (SRAM-CIMs) have attracted widespread attention owing to its good scalability with advanced process. At present, a rich variety of works focus on energy-efficiency improvement by either designing different bit-cell structures or optimizing circuit/chip architectures. However, owing to the CIM's primitive property to store one of the operands in the memory bit-cells, substantial computing resource is wasted by suspension during the operands loading procedure. In this paper, a high-throughput SRAM-CIM (HiT-CIM) architecture with simultaneous weight loading and computing capabilities is proposed by integrating on-chip nonvolatile MRAM (magnetic random-access memory). Meanwhile, both the mainstream current-domain and charge-domain SRAM bit-cell structures are optimized to support such an architecture. Furthermore, a reconfigurable fully-pipelined MRAM is designed to provide fast data loading in HiT-CIM, which can finetune weight loading strategy rapidly for different neural network models. Afterwards, an optimal evaluation and configuration strategy is proposed to improve the macro-level performance by considering the key components and parameters in terms of SRAM array, ADC, MRAM structure and frequency. Finally, the HiT-CIM's feasibility is verified under a 40-nm foundry's process. The results show that a multiple-fold speed improvement can be obtained on VGG19, ResNet18 and MobileNetV1, respectively. In specific, the area efficiency of HiT-CIM on VGG19 achieves 1124 GOPS/mm2 and 1880.12 GOPS/mm2 for the current-domain and chargedomain SRAM-CIMs, respectively. Up to 5.3× improvement is realized compared with prior worksKeywords
Funding Information
- National Natural Science Foundation of China (62274008)
- Beijing Natural Science Foundation (L223004)
This publication has 39 references indexed in Scilit:
- PipeFuser: Building Flexible Pipeline Architecture for DNN Accelerators via Layer Fusion2024
- In‐memory calculation with embedded arithmetic and logic units for deep neural networkElectronics Letters, 2022
- ShareFloat CIM: A Compute-In-Memory Architecture with Floating-Point Multiply-and-Accumulate Operations2022
- A crossbar array of magnetoresistive memory devices for in-memory computingNature, 2022
- Field-free switching of a perpendicular magnetic tunnel junction through the interplay of spin–orbit and spin-transfer torquesNature Electronics, 2018
- Deep Continuous Fusion for Multi-sensor 3D Object Detection2018
- Current-induced magnetization switching in atom-thick tungsten engineered perpendicular magnetic tunnel junctions with large tunnel magnetoresistanceNature Communications, 2018
- 28-nm 1T-1MTJ 8Mb 64 I/O STT-MRAM with symmetric 3-section reference structure and cross-coupled sensing amplifier2017
- Deep learningNature, 2015
- Minimizing Computation in Convolutional Neural Networks2014