Metraipolis › archive › dc_cc5efa377ede

RecIS: Unifying Sparse and Dense Computation for Efficient Recommendation Training — arXiv:2610.41457

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.41457
License (as archived)
CC0-1.0
Archived
2026-11-02T07:05:44.905Z
SHA-256 of the archived text
f151f710434bf6a0822b94ad3ae5d32220df18cb973945d056aa5b86ffe1f211
Stable URL of this page
https://metraipolis.com/documento/dc_cc5efa377ede

How to cite

RecIS: Unifying Sparse and Dense Computation for Efficient Recommendation Training — arXiv:2610.41457 — arXiv (en), https://arxiv.org/abs/2610.41457 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_cc5efa377ede (Metraipolis archive dc_cc5efa377ede, sha256 f151f710434bf6a0822b94ad3ae5d32220df18cb973945d056aa5b86ffe1f211)
BibTeX
@misc{dccc5efa377ede,
  title = {RecIS: Unifying Sparse and Dense Computation for Efficient Recommendation Training — arXiv:2610.41457},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.41457},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_cc5efa377ede; sha256 f151f710434bf6a0822b94ad3ae5d32220df18cb973945d056aa5b86ffe1f211},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.41457. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Modern recommender models scale sparse representations and dense computation together. They process thousands of heterogeneous feature fields, long user-behavior sequences, and trillion-scale Embedding parameters, while their dense networks increasingly use large Transformer architectures. The resulting execution mismatch creates fine-grained operations within stages, load imbalance and synchronization delays across devices, and GPU resource bubbles between stages. As models grow, these costs become a primary limit to scalable recommender training. Although existing systems improve Embedding caching and data prefetching, these inefficiencies still coexist in the GPU-centric workloads we study and jointly limit end-to-end throughput. We present RecIS, a unified framework for GPU-centric recommender training. RecIS treats the complete path from input features to dense computation as one optimization target and coordinates three complementary mechanisms. First, adaptive fusion combines fragmented feature transformations with vectorized memory access. Second, a GPU Embedding engine uses hash-based sharding and hierarchical communication to balance request placement and eliminate redundant transfers; it also specializes aggregation for forward and backward index structures. Third, a multistage pipeline overlaps input preparation for future iterations with dense computation in the current iteration. Together, these mechanisms reorganize a fragmented sparse front end into a dataflow that continuously feeds dense computation. We evaluate RecIS on representative industrial recommender models across distributed training platforms. Compared with PICASSO and TorchRec, RecIS reaches up to $2.62\times$ and $1.15\times$ their throughput, respectively, while reducing step latency by up to 61.5\% and 12.9\%. RecIS provides an efficient and scalable training system for large recommender models.

Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.DC, nuovi lavori del giorno).
Categorie arXiv: Distributed, Parallel, and Cluster Computing (cs.DC).
Pagina del lavoro: https://arxiv.org/abs/2610.41457 — 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 2026-11-02T06:57:28.255Z dalla pagina https://arxiv.org/list/cs.DC/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).