TY - GEN
T1 - Cost-Performance Analysis
T2 - 26th International Conference on Parallel and Distributed Computing, Applications and Technologies, PDCAT 2025
AU - Barrak, Amine
AU - Petrillo, Fabio
AU - Jaafar, Fehmi
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2026.
PY - 2026
Y1 - 2026
N2 - This paper presents a comparative evaluation of four serverless training frameworks: SPIRT, MLLess, LambdaML AllReduce, and ScatterReduce, alongside a GPU-based baseline, using CNN models on CIFAR-10. We assess each architecture across training time, cost, communication overhead, and accuracy under consistent experimental conditions. While GPU-based training achieves the fastest convergence and highest accuracy, serverless frameworks offer cost advantages for lightweight models. Optimizations such as gradient accumulation and in-database computation improve serverless performance. Our findings reveal key trade-offs and highlight the potential of GPU-backed serverless platforms for scalable distributed training.
AB - This paper presents a comparative evaluation of four serverless training frameworks: SPIRT, MLLess, LambdaML AllReduce, and ScatterReduce, alongside a GPU-based baseline, using CNN models on CIFAR-10. We assess each architecture across training time, cost, communication overhead, and accuracy under consistent experimental conditions. While GPU-based training achieves the fastest convergence and highest accuracy, serverless frameworks offer cost advantages for lightweight models. Optimizations such as gradient accumulation and in-database computation improve serverless performance. Our findings reveal key trade-offs and highlight the potential of GPU-backed serverless platforms for scalable distributed training.
KW - Cost-effectiveness
KW - Distributed Machine Learning
KW - Serverless ML Architectures
UR - https://www.scopus.com/pages/publications/105048243182
U2 - 10.1007/978-981-95-9846-5_31
DO - 10.1007/978-981-95-9846-5_31
M3 - Contribution to conference proceedings
AN - SCOPUS:105048243182
SN - 9789819598458
T3 - Lecture Notes in Computer Science
SP - 368
EP - 380
BT - Parallel and Distributed Computing, Applications and Technologies - 26th International Conference, PDCAT 2025, Proceedings
A2 - Tian, Hui
A2 - Park, James Jong Hyuk
A2 - Zhang, Yong
PB - Springer Science and Business Media Deutschland GmbH
Y2 - 22 November 2025 through 24 November 2025
ER -