EUVIMEDEUVIMED entdecken
Crossref

THERMOS: Thermally-Aware Multi-Objective Scheduling of AI Workloads on Heterogeneous Multi-Chiplet PIM Architectures

Alish Kanani, Lukas Pfromm, Harsh Sharma, Jana Doppa, Partha Pande, Umit Ogras

ACM Transactions on Embedded Computing Systems · 30.11.2025 · 24 (5s) · 1-26

DOI: 10.1145/3762655 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ACM Transactions on Embedded Computing Systems, online erschienen am 26.09.2025. Verfasst von Alish Kanani und 5 weiteren Autor:innen. Die Arbeit wurde bisher 9-mal zitiert (Crossref) und nennt 73 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Chiplet-based integration enables large-scale systems that combine diverse technologies, enabling higher yield, lower costs, and scalability, making them well-suited to AI workloads. Processing-in-Memory (PIM) has emerged as a promising solution for AI inference, leveraging technologies such as ReRAM, SRAM, and FeFET, each offering unique advantages and tradeoffs. A heterogeneous chiplet-based PIM architecture can harness the complementary strengths of these technologies to enable higher performance and energy efficiency. However, scheduling AI workloads across such a heterogeneous system is challenging due to competing performance objectives, dynamic workload characteristics, and power and thermal constraints. To address this need, we propose THERMOS, a thermally-aware, multi-objective scheduling framework for AI workloads on heterogeneous multi-chiplet PIM architectures. THERMOS trains a single multi-objective reinforcement learning (MORL) policy that is capable of achieving Pareto-optimal execution time, energy, or a balanced objective at runtime, depending on the target preferences. Comprehensive evaluations show that THERMOS achieves up to 89% faster average execution time and 57% lower average energy consumption than baseline AI workload scheduling algorithms with only 0.14% runtime and 0.022% energy overhead.

Zitieren

APA

Kanani, A., Pfromm, L., Sharma, H., Doppa, J., Pande, P., & Ogras, U. (2025). THERMOS: Thermally-Aware Multi-Objective Scheduling of AI Workloads on Heterogeneous Multi-Chiplet PIM Architectures. ACM Transactions on Embedded Computing Systems, 24(5s), 1-26. https://doi.org/10.1145/3762655

Vancouver

Kanani A, Pfromm L, Sharma H, Doppa J, Pande P, Ogras U. THERMOS: Thermally-Aware Multi-Objective Scheduling of AI Workloads on Heterogeneous Multi-Chiplet PIM Architectures. ACM Transactions on Embedded Computing Systems. 2025;24(5s):1-26. doi:10.1145/3762655

BibTeX
@article{Kanani2025,
  author = {Kanani, Alish and Pfromm, Lukas and Sharma, Harsh and Doppa, Jana and Pande, Partha and Ogras, Umit},
  title = {THERMOS: Thermally-Aware Multi-Objective Scheduling of AI Workloads on Heterogeneous Multi-Chiplet PIM Architectures},
  journal = {ACM Transactions on Embedded Computing Systems},
  year = {2025},
  volume = {24},
  number = {5s},
  pages = {1--26},
  doi = {10.1145/3762655}
}
Kostenlos starten →