Metraipolis › archive › dc_ddde31de54d0

Deep Learning Compilers: Architecture, Optimization, and Empirical Characterization — arXiv:2611.04273

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.04273
License (as archived)
CC0-1.0
Archived
2026-11-04T21:46:14.773Z
SHA-256 of the archived text
c1067b85b3de25eb823e29877cca123cdb711261e8ba318ec4f215cb6b022241
Stable URL of this page
https://metraipolis.com/documento/dc_ddde31de54d0

How to cite

Deep Learning Compilers: Architecture, Optimization, and Empirical Characterization — arXiv:2611.04273 — arXiv (en), https://arxiv.org/abs/2611.04273 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ddde31de54d0 (Metraipolis archive dc_ddde31de54d0, sha256 c1067b85b3de25eb823e29877cca123cdb711261e8ba318ec4f215cb6b022241)
BibTeX
@misc{dcddde31de54d0,
  title = {Deep Learning Compilers: Architecture, Optimization, and Empirical Characterization — arXiv:2611.04273},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.04273},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ddde31de54d0; sha256 c1067b85b3de25eb823e29877cca123cdb711261e8ba318ec4f215cb6b022241},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.04273. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

As deep learning (DL) models become increasingly complex and hardware architectures grow more heterogeneous, DL compilers play a critical role in optimizing computational efficiency and enabling scalable deployment. This paper presents a systematic analysis of DL compilers, covering their architectural evolution, compilation workflows, optimization techniques, and emerging research directions. We organize DL compiler architectures into frontend, middle-end, and backend stages, emphasizing the role of the middle-end in connecting high-level model representations with hardware-specific optimizations. We further examine representative DL compilers and recent optimization techniques across different compilation stages. Beyond architectural and technical analysis, we conduct an empirical evaluation of compiler behaviors to investigate the effectiveness of graph optimization, compilation overhead, and dynamic-shape handling. Experiments using PyTorch/TorchInductor on an NVIDIA RTX A6000 demonstrate that compiler benefits depend strongly on workload characteristics, with speedups of up to 8.21\times for fusion-friendly computation graphs, while other workloads exhibit limited improvements. Our results further reveal that compilation overhead can require thousands of executions to amortize, and dynamic-shape compilation reduces cumulative shape-transition costs by approximately 70\% at the expense of modest steady-state performance degradation. By combining systematic architectural analysis with quantitative experimental evidence, this work provides insights into the design, optimization, and practical deployment of modern DL compilers.

Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.AR, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Hardware Architecture (cs.AR).
Pagina del lavoro: https://arxiv.org/abs/2611.04273 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-04T21:36:34.399Z dalla pagina https://arxiv.org/list/cs.AR/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).