Communication Dans Un Congrès Année : 2025

Systolic Arrays and Structured Pruning Co-design for Efficient Transformers in Edge Systems

Résumé

Efficient deployment of resource-intensive transformers on edge devices necessitates cross-stack optimization. We thus study the interrelation between structured pruning and systolic acceleration, matching the size of pruned blocks with the systolic array dimensions. In this setting, computations of pruned weight blocks can be skipped, reducing run-time and energy consumption, but potentially impacting quality of service (QoS). To evaluate the trade-offs between systolic array size and sparsity opportunities, we present a novel co-design framework that integrates algorithmic optimization, system simulation, and hardware design. Targeting speech recognition and machine translation using transformers as case study, we analyze how configuration choices across the stack affect performance metrics. Results demonstrate that structured pruning on systems featuring systolic array acceleration can effectively increase performance, while maintaining high QoS levels. Up to 44% system-wide speedups due to structured pruning and quantization were measured, with only 1.4% word error rate degradation on the standard LibriSpeech dataset.

CCS Concepts

• Hardware → Hardware-software codesign; • Computer systems organization → Systolic arrays; • Computing methodologies → Neural networks.

Fichier principal
Vignette du fichier
GLSVLSI_2025_Pedro.pdf (1.43 Mo) Télécharger le fichier

Dates et versions

hal-05242392 , version 1 (05-09-2025)

Licence

Identifiants

Citer

Pedro Palacios, Rafael Medina, Jean-Luc Rouas, Giovanni Ansaloni, David Atienza. Systolic Arrays and Structured Pruning Co-design for Efficient Transformers in Edge Systems. GLSVLSI '25: Great Lakes Symposium on VLSI 2025, Apr 2025, New Orleans LA USA, United States. pp.320-327, ⟨10.1145/3716368.3735158⟩. ⟨hal-05242392⟩
66 Consultations
131 Téléchargements

Altmetric

Partager

  • More