EUVIMEDEUVIMED entdecken
Crossref

BF-IMNA: A Bit Fluid In-Memory Neural Architecture for Neural Network Acceleration

Mariam Rakka, Rachid Karami, Ahmed M. Eltawil, Mohammed E. Fouda, Fadi Kurdahi

ACM Journal on Emerging Technologies in Computing Systems · 11.11.2026

DOI: 10.1145/3860531 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ACM Journal on Emerging Technologies in Computing Systems, online erschienen am 11.11.2026. Verfasst von Mariam Rakka und 4 weiteren Autor:innen. Die Arbeit nennt 42 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Mixed-precision quantization, which assigns precisions per-layer, per-network, per-channel, or per-parameter, is gaining traction for efficient Neural Network (NN) realization with higher throughput and lower energy. Unlike conventional fixed-precision accelerators, BF-IMNA advances the mixed-precision design space through an AP-based in-memory implementation. We present BF-IMNA, a bit fluid IMC accelerator for end-to-end Convolutional NN inference supporting static and dynamic mixed-precision without run-time reconfiguration overhead. At its heart are Associative Processors (APs), bit-serial word-parallel SIMD-like engines. We evaluate ImageNet inference with AlexNet, VGG16, and ResNet50 across SRAM and ReRAM CAM cells, mixed-precision configurations, and supply voltages. To demonstrate bit fluidity, we apply HAWQ-V3's published ResNet18 schedules to the BF-IMNA cost model and pair them with HAWQ-V3's published accuracies and model sizes. The results reveal an accuracy–Energy-Delay Product trade-off. The high-latency point is \(1.16\) points below INT8 accuracy with \(1.13\times\) better EDP, while the low-latency point has \(1.91\times\) better EDP and a \(45.5\%\) smaller model at a \(3.00\) -point accuracy reduction. At 500-MHz, true-batch-one comparison, BitFusion provides lower absolute energy and latency, while BF-IMNA provides lower average power at all five operating points, \(2106\times\) greater on-chip weight capacity ( \(32.8\times\) per unit modeled area), and larger relative energy reductions at aggressive INT4 and low-latency HAWQ points.

Zitieren

APA

Rakka, M., Karami, R., Eltawil, A. M., Fouda, M. E., & Kurdahi, F. (2026). BF-IMNA: A Bit Fluid In-Memory Neural Architecture for Neural Network Acceleration. ACM Journal on Emerging Technologies in Computing Systems. https://doi.org/10.1145/3860531

Vancouver

Rakka M, Karami R, Eltawil AM, Fouda ME, Kurdahi F. BF-IMNA: A Bit Fluid In-Memory Neural Architecture for Neural Network Acceleration. ACM Journal on Emerging Technologies in Computing Systems. 2026. doi:10.1145/3860531

BibTeX
@article{Rakka2026,
  author = {Rakka, Mariam and Karami, Rachid and Eltawil, Ahmed M. and Fouda, Mohammed E. and Kurdahi, Fadi},
  title = {BF-IMNA: A Bit Fluid In-Memory Neural Architecture for Neural Network Acceleration},
  journal = {ACM Journal on Emerging Technologies in Computing Systems},
  year = {2026},
  doi = {10.1145/3860531}
}
Kostenlos starten →