English

Why Smaller Is Slower? Dimensional Misalignment in Compressed LLMs

Distributed, Parallel, and Cluster Computing 2026-04-14 v1 Artificial Intelligence

Abstract

Post-training compression reduces LLM parameter counts but often produces irregular tensor dimensions that degrade GPU performance -- a phenomenon we call \emph{dimensional misalignment}. We present a full-stack analysis tracing root causes at three levels: framework, library, and hardware. The key insight is that model inference becomes slower because the resulting dimensions are unfriendly with the GPU execution stack. For example, compressing Llama-3-8B with activation-aware singular value decomposition (ASVD) has 15\% fewer parameters yet runs no faster than the uncompressed baseline, because 95\% of its dimensions are misaligned. We propose \textbf{GAC} (GPU-Aligned Compression), a new compression paradigm that wraps any dimension-reducing compressor and re-selects hardware-aligned dimensions via multi-choice knapsack optimization under the same parameter budget. We evaluate GAC on Llama-3-8B with ASVD and LLM-Pruner, achieving 100\% alignment and recovering up to 1.5×\times speedup while preserving model quality.

Keywords

Cite

@article{arxiv.2604.09595,
  title  = {Why Smaller Is Slower? Dimensional Misalignment in Compressed LLMs},
  author = {Jihao Xin and Tian Lyu and Qilong Pan and Kesen Wang and Marco Canini},
  journal= {arXiv preprint arXiv:2604.09595},
  year   = {2026}
}
R2 v1 2026-07-01T12:03:20.558Z