NestPipe achieves up to 3.06x speedup and 94.07% scaling efficiency on 1,536 workers via dual-buffer inter-batch and frozen-window intra-batch pipelining that overlaps communication with computation.
Two-dimensional sparse parallelism for large scale deep learning recommendation model training
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
baseline 1
citation-polarity summary
fields
cs.DC 2years
2026 2verdicts
UNVERDICTED 2roles
baseline 1polarities
baseline 1representative citing papers
TurboGR trains up to 0.2B-parameter generative recommendation models on Ascend NPUs at 54.71% MFU with 0.97 near-linear scalability via jagged acceleration, hierarchical parallelism, and negative sampling optimizations.
citing papers explorer
-
NestPipe: Large-Scale Recommendation Training on 1,500+ Accelerators via Nested Pipelining
NestPipe achieves up to 3.06x speedup and 94.07% scaling efficiency on 1,536 workers via dual-buffer inter-batch and frozen-window intra-batch pipelining that overlaps communication with computation.
-
TurboGR: An Accelerated Training System for Large-Scale Generative Recommendation
TurboGR trains up to 0.2B-parameter generative recommendation models on Ascend NPUs at 54.71% MFU with 0.97 near-linear scalability via jagged acceleration, hierarchical parallelism, and negative sampling optimizations.