SwiftQK replaces the All-Gather used for layerwise QK-Norm under tensor parallelism with scalar sum-of-squares exchange, overlapped with computation in a fused persistent kernel, cutting QK-Norm latency by 81-94%.
Flashoverlap: A lightweight design for efficiently overlapping communication and computation,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
SwiftQK: Fast and Communication-Efficient Tensor Parallelism for Query-Key Normalization
SwiftQK replaces the All-Gather used for layerwise QK-Norm under tensor parallelism with scalar sum-of-squares exchange, overlapped with computation in a fused persistent kernel, cutting QK-Norm latency by 81-94%.