Logo des Repositoriums
 
Zeitschriftenartikel

Comparing GPU and TPU in an Iterative Scenario: A Study on Neural Network-based Image Generation

Vorschaubild nicht verfügbar

Volltext URI

Dokumententyp

Text/Journal Article

Zusatzinformation

Datum

2024

Zeitschriftentitel

ISSN der Zeitschrift

Bandtitel

Verlag

Gesellschaft für Informatik e.V., Fachgruppe PARS

Zusammenfassung

This paper explores the utilization of TPUs (Tensor Processing Units) and GPUs (Graphics Processing Units) in iterative applications involving neural networks. We employ a Pix2Pix approachfor computing sequential flows, evaluating the effectiveness in scenarios where NNs are only a component of the system. While TPUs demonstrate performance improvements during training with large batch sizes, we observe no significant acceleration during inference compared to GPUs. The study highlights the need to carefully consider workload and system architecture when incorporating TPUs, emphasizing that their advantages are more prominent in training tasks.

Beschreibung

Lehmann, Roman; Schaarschmidt, Paul; Karl, Wolfgang (2024): Comparing GPU and TPU in an Iterative Scenario: A Study on Neural Network-based Image Generation. PARS-Mitteilungen: Vol. 36. Gesellschaft für Informatik e.V., Fachgruppe PARS. ISSN: 0177-0454

Schlagwörter

Zitierform

DOI

Tags