Skip to main content
Submit
ProceedingsShort paper Open access

Profiling Energy Use of Small Language Models on Mobile SoCs

In Proceedings of the Symposium on Efficient and On-Device Machine Learning (SEODML 2026), pp. 21–26. Singapore, 8 Sep 2026.

DOI 10.xxxx/ctm.pr.2026.0008 · ID CTM-PR-2026-0008

Download PDF

Cite this record

Saito, Y. & Tanaka, M. (2026). Profiling Energy Use of Small Language Models on Mobile SoCs. In Proceedings of the Symposium on Efficient and On-Device Machine Learning (SEODML 2026) (pp. 21–26). CONTINUUM Proceedings. doi:10.xxxx/ctm.pr.2026.0008

Sample record. Fictional demonstration content.

Abstract

We measure energy per generated token for eleven small language models on three mobile systems-on-chip. Memory traffic, not arithmetic, explains most of the variation, and the best configuration per device differs by up to 2.4×.

Keywords energy · mobile · profiling

Full text

CTM-PR-2026-0008.pdf

Paper · PDF · pp. 21–26

Download

Demonstration record: no file is attached.

How to cite

Saito, Y. & Tanaka, M. (2026). Profiling Energy Use of Small Language Models on Mobile SoCs. In Proceedings of the Symposium on Efficient and On-Device Machine Learning (SEODML 2026) (pp. 21–26). CONTINUUM Proceedings. doi:10.xxxx/ctm.pr.2026.0008

More from SEODML 2026