EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A High-Precision Hybrid Floating-Point Compute-in-Memory Architecture for Complex Deep Learning

Zizhao Ma, Chunshan Wang, Qi Chen, Yifan Wang, Yufeng Xie

Electronics · 13.11.2025 · 14 (22) · 4414

DOI: 10.3390/electronics14224414 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Electronics, online erschienen am 13.11.2025. Verfasst von Zizhao Ma und 4 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 38 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

As artificial intelligence (AI) advances, deep learning models are shifting from convolutional architectures to transformer-based structures, highlighting the importance of accurate floating-point (FP) calculations. Compute-in-memory (CIM) enhances matrix multiplication performance by breaking down the von Neumann architecture. However, many FPCIMs struggle to maintain high precision while achieving efficiency. This work proposes a high-precision hybrid floating-point compute-in-memory (Hy-FPCIM) architecture for Vision Transformer (ViT) through post-alignment with two different CIM macros: Bit-wise Exponent Macro (BEM) and Booth Mantissa Macro (BMM). The high-parallelism BEM efficiently implements exponent calculations in-memory with the Bit-Separated Exponent Summation Unit (BSESU) and the routing-efficient Bit-wise Max Finder (BMF). The high-precision BMM achieves nearly lossless mantissa computation in-memory with efficient Booth 4 encoding and the sensitivity-amplifier-free Flying Mantissa Lookup Table based on 12T Triple Port SRAM. The proposed Hy-FPCIM architecture achieves 23.7 TFLOPS/W energy efficiency and 0.754 TFLOPS/mm2 area efficiency, with 617 Kb/mm2 memory density in 28 nm technology. With almost lossless architectures, the proposed Hy-FPCIM achieves an accuracy of 81.04% in recognition tasks on the ImageNet dataset using ViT, representing a 0.03% decrease compared to the software baseline. This research presents significant advantages in both accuracy and energy efficiency, providing critical technology for complex deep learning applications.

Zitieren

APA

Ma, Z., Wang, C., Chen, Q., Wang, Y., & Xie, Y. (2025). A High-Precision Hybrid Floating-Point Compute-in-Memory Architecture for Complex Deep Learning. Electronics, 14(22), 4414. https://doi.org/10.3390/electronics14224414

Vancouver

Ma Z, Wang C, Chen Q, Wang Y, Xie Y. A High-Precision Hybrid Floating-Point Compute-in-Memory Architecture for Complex Deep Learning. Electronics. 2025;14(22):4414. doi:10.3390/electronics14224414

BibTeX
@article{Ma2025,
  author = {Ma, Zizhao and Wang, Chunshan and Chen, Qi and Wang, Yifan and Xie, Yufeng},
  title = {A High-Precision Hybrid Floating-Point Compute-in-Memory Architecture for Complex Deep Learning},
  journal = {Electronics},
  year = {2025},
  volume = {14},
  number = {22},
  pages = {4414},
  doi = {10.3390/electronics14224414}
}
Kostenlos starten →