cfaed Publications

Compiler-Assisted Speculative Sampling for Accelerated LLM Inference on Heterogeneous Edge Devices

Reference

Alejandro Ruiz y Mesa, Guilherme Korol, Moritz Riesteter, João Paulo Cardoso de Lima, Jeronimo Castrillon, "Compiler-Assisted Speculative Sampling for Accelerated LLM Inference on Heterogeneous Edge Devices", In Proceeding: 8th Workshop on Accelerated Machine Learning (AccML), co-located with 21st International Conference on High-Performance and Embedded Architectures and Compilers (HiPEAC), 8pp, Jan 2026.

Bibtex

@InProceedings{ruiz_y_mesa_accml26,
author = {Alejandro Ruiz y Mesa and Guilherme Korol and Moritz Riesteter and João Paulo Cardoso de Lima and Jeronimo Castrillon},
booktitle = {8th Workshop on Accelerated Machine Learning (AccML), co-located with 21st International Conference on High-Performance and Embedded Architectures and Compilers (HiPEAC)},
title = {Compiler-Assisted Speculative Sampling for Accelerated LLM Inference on Heterogeneous Edge Devices},
location = {Kraków, Poland},
pages = {8pp},
month = jan,
year = {2026},
url = {https://accml.dcs.gla.ac.uk/papers/2026/8th_AccML_paper_8.pdf},
}

Downloads

2601_Ruiz_AccML [PDF]

Permalink

https://cfaed.tu-dresden.de/publications?pubId=3873


Go back to publications list