Metraipolis › archive › dc_08df7e1fcdc7

PairSight: Characterizing and Predicting Operator-Level Interference in GPU Sharing for DNN Inference — arXiv:2610.23086

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.23086
License (as archived)
CC0-1.0
Archived
2026-10-19T08:44:24.506Z
SHA-256 of the archived text
61cfc3f2a6a71bab05910e63769417395b2d7eadd5d0f6c9a03619ad39eaa05f
Stable URL of this page
https://metraipolis.com/documento/dc_08df7e1fcdc7

How to cite

PairSight: Characterizing and Predicting Operator-Level Interference in GPU Sharing for DNN Inference — arXiv:2610.23086 — arXiv (en), https://arxiv.org/abs/2610.23086 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_08df7e1fcdc7 (Metraipolis archive dc_08df7e1fcdc7, sha256 61cfc3f2a6a71bab05910e63769417395b2d7eadd5d0f6c9a03619ad39eaa05f)
BibTeX
@misc{dc08df7e1fcdc7,
  title = {PairSight: Characterizing and Predicting Operator-Level Interference in GPU Sharing for DNN Inference — arXiv:2610.23086},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.23086},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_08df7e1fcdc7; sha256 61cfc3f2a6a71bab05910e63769417395b2d7eadd5d0f6c9a03619ad39eaa05f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.23086. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Serving DNN inference is costly, yet a single task seldom saturates both the compute units and the memory bandwidth of a GPU. GPU sharing raises utilization by co-locating several tasks on one GPU, at the cost of interference among them. Existing systems estimate this interference with binary compute-memory rules, standalone scaling curves, or model-level predictors. However, these estimates rest on a limited understanding of interference. They leave the software and hardware properties of co-located operators largely unexplored and give no slowdown prediction that is both interpretable and accurate. We conduct an operator-level study of interference under SM partitioning and use hardware counters to trace it to contention on the shared memory path, where the slowdown depends jointly on both co-located operators. Based on these findings, we propose PairSight, a hardware-aware interference predictor that learns how the hardware behaviors of co-located tasks interact and predicts the latency of each task before co-location. On six DNN models including LLMs, PairSight predicts latency under sharing with 4.3% mean error, against 19.0% for a compute-memory intensity rule, and adapts to unseen models and inputs with a few samples.

Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.PF, nuovi lavori del giorno).
Categorie arXiv: Performance (cs.PF).
Pagina del lavoro: https://arxiv.org/abs/2610.23086 — 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 2026-10-19T08:40:09.816Z dalla pagina https://arxiv.org/list/cs.PF/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).