EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Bridging Architectures, Mapping, and Learning for DNN Acceleration with Processing-in-Memory and In-Memory Computing Systems

Syeda Munazza Marium, Song Chen

Microelectronics · 10.06.2026 · 2 (2) · 10

DOI: 10.3390/microelectronics2020010 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Microelectronics, online erschienen am 10.06.2026. Verfasst von Syeda Munazza Marium und 1 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 268 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Processing-in-memory and in-memory computing (PIM/IMC) are increasingly explored to mitigate the von Neumann data-movement bottleneck that limits deep neural network (DNN) performance and energy efficiency. Progress, however, remains fragmented across device substrates, architectural prototypes, mapping and scheduling methods, compiler toolchains, and benchmarking practices, making results hard to compare and slowing deployment. This survey synthesizes developments from 2019–2025 along four coupled axes: (i) memory substrates and architectural design, (ii) mapping, partitioning, and scheduling, including learning- and graph-based strategies, (iii) compilers and end-to-end deployment flows, and (iv) benchmarking datasets, metrics, and reporting norms. Drawing on over twenty representative platforms spanning static random-access memory (SRAM) and dynamic random-access memory (DRAM), emerging non-volatile, capacitive, and photonic substrates, we clarify the trade-offs separating analog/charge-domain IMC from digital SRAM/DRAM-centric PIM, including reported peaks up to 600 TOPS/W and 1.5 TOPS/mm2. We organize mapping frameworks into a unified reference taxonomy, identify recurrent evaluation pitfalls that undermine reproducibility, and highlight persistent gaps in training support, robustness under non-idealities, and coverage of large-scale GNN workloads. Finally, we outline a five-phase roadmap from benchmark standardization to industrial validation toward compiler-integrated, GNN-informed PIM/IMC systems validated on production-scale workloads.

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Marium, S. M., & Chen, S. (2026). Bridging Architectures, Mapping, and Learning for DNN Acceleration with Processing-in-Memory and In-Memory Computing Systems. Microelectronics, 2(2), 10. https://doi.org/10.3390/microelectronics2020010

Vancouver

Marium SM, Chen S. Bridging Architectures, Mapping, and Learning for DNN Acceleration with Processing-in-Memory and In-Memory Computing Systems. Microelectronics. 2026;2(2):10. doi:10.3390/microelectronics2020010

BibTeX
@article{Marium2026,
  author = {Marium, Syeda Munazza and Chen, Song},
  title = {Bridging Architectures, Mapping, and Learning for DNN Acceleration with Processing-in-Memory and In-Memory Computing Systems},
  journal = {Microelectronics},
  year = {2026},
  volume = {2},
  number = {2},
  pages = {10},
  doi = {10.3390/microelectronics2020010}
}
Kostenlos starten →