floating point (FP)

  • No Comments

A Nonvolatile AI-Edge Processor With Lossless-Compressed-Computing STT-MRAM Near-Memory-Compute Macro Using Dynamic Floating-/Fixed-Point Accumulation

A Nonvolatile AI-Edge Processor With Lossless-Compressed-Computing STT-MRAM Near-Memory-Compute Macro Using Dynamic Floating-/Fixed-Point Accumulation 150 150
Author(s): De-Qi You, Win-San Khwa, Bo Zhang, Jui-Jen Wu, Chuan-Jia Jhang, Guan-Yi Lin, Po-Jung Chen, Ting-Chien Chiu, Fang-Yi Chen, Andrew Lee, Yu-Cheng Hung, Chung-Chuan Lo, Ren-Shuo Liu, Chih-Cheng Hsieh, Kea-Tiong Tang, Yu-Der Chih, Tsung-Yung Jonathan Chang, Meng-Fan Chang

Abstract:

Nonvolatile AI-edge processors based on near-memory-compute (nvNMC) enable energy-efficient multiply-and-accumulate (MAC) operations with short wakeup latency for edge inference operations. Lossless compression is required for floating-point (FP) neural network (NN) models under on-chip memory capacity constraints; however, this imposes several challenges: 1) data decompression overhead due to lossless FP weight encoding; 2) …

Published in: IEEE Journal of Solid-State Circuits
Page(s): 1 – 13
Year of Publication: 2026
Electronic ISSN: 1558-173X
DOI: 10.1109/JSSC.2026.3705432
Publisher: IEEE

View on IEEE Xplore

  • No Comments

A 3 nm FinFET 125 TOPS/W-29 TFLOPS/W, 90 TOPS/mm2-17 TFLOPS/mm2 SRAM-Based INT8, and FP16 Digital-CIM Compiler With Support for Multi-Weight Update/Cycle

A 3 nm FinFET 125 TOPS/W-29 TFLOPS/W, 90 TOPS/mm2-17 TFLOPS/mm2 SRAM-Based INT8, and FP16 Digital-CIM Compiler With Support for Multi-Weight Update/Cycle 150 150
Author(s): Haruki Mori, Je-Min Hung, Wei-Chang Zhao, Kinshuk Khare, Brian Crafton, Hiroaki Ishii, Cheng-En Lee, Xiaochen Peng, Xiaoyu Sun, Yorinobu Fujino, Chia-Jung Tsen, Vineet Joshi, Chao-Kai Chuang, Takeshi Hashizume, Chia-Fu Lee, Tan-Li Chou, Kerem Akarvardar, Saman Adham, Yih Wang, Hidehiro Fujiwara, Yu-Der Chih, Yen-Huei Chen, Hung-Jen Liao, Tsung-Yung Jonathan Chang

Abstract:

This article presents an static random-access memory (SRAM)-based digital compute-in-memory (CIM) compiler implemented with 3 nm high- $\kappa $ metal gate (HKMG) FinFET technology, supporting flexible INT8 and FP16 formats for weight and activation multiply-accumulate (MAC) operations, offering configuration flexibility, high accuracy, and improved area and power efficiency. The FP16 digital …

Published in: IEEE Journal of Solid-State Circuits
Page(s): 1502 – 1512
Year of Publication: 2026
Electronic ISSN: 1558-173X
DOI: 10.1109/JSSC.2025.3646072
Publisher: IEEE

View on IEEE Xplore

  • No Comments

Denim: Heterogeneous Compute-in-Memory Accelerator Exploiting Denoising–Similarity for Diffusion Models

Denim: Heterogeneous Compute-in-Memory Accelerator Exploiting Denoising–Similarity for Diffusion Models 150 150
Author(s): Ruiqi Guo, Lei Wang, Xiaofeng Chen, Keyi Jiang, Zhiheng Yue, Huiming Han, Yang Wang, Fengbin Tu, Shaojun Wei, Yang Hu, Shouyi Yin

Abstract:

Diffusion models have recently revolutionized the field of image synthesis due to their ability to generate photorealistic images. However, one of the main drawbacks of diffusion models is that the image generation process is expensive. Large image-to-image networks have to be applied multiple times in order to iteratively optimize the …

Published in: IEEE Journal of Solid-State Circuits
Page(s): 1 – 15
Year of Publication: 2026
Electronic ISSN: 1558-173X
DOI: 10.1109/JSSC.2026.3684745
Publisher: IEEE

View on IEEE Xplore

  • No Comments

A 28-nm Digital Transpose SRAM Compute-in-Memory Macro With Accurate/Approximate Dual Mode for Floating-Point Edge Training and Inference

A 28-nm Digital Transpose SRAM Compute-in-Memory Macro With Accurate/Approximate Dual Mode for Floating-Point Edge Training and Inference 150 150
Author(s): Yiyang Yuan, Bingxin Zhang, Yiming Yang, Yishan Luo, Qirui Chen, Haitao Wang, Qihao Liu, Zhiming Chen, Hao Wu, Jinshan Yue, Shidong Lv, Xinghua Wang, Pui-In Mak, Xiaoran Li, Feng Zhang

Abstract:

Static random-access memory (SRAM)-based computing-in-memory (CIM) macros have been widely studied to improve the energy efficiency of edge artificial intelligence (AI) inference tasks. However, less attention has been given to AI training, which requires CIM macros to not only perform matrix multiply-accumulate (MAC) operations but also support matrix transposition. …

Published in: IEEE Journal of Solid-State Circuits
Page(s): 1 – 13
Year of Publication: 2026
Electronic ISSN: 1558-173X
DOI: 10.1109/JSSC.2026.3679560
Publisher: IEEE

View on IEEE Xplore

Search this website