跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.21974 · v1

Beyond Bitwidth: Deployment-Constrained Mixed-Precision Quantization for Multilingual LLM Inference on NVIDIA Blackwell

首次提交 2026-11-17 06:58
版本更新 2026-11-17 06:58
雷达观测 2026-11-18 19:43
arXiv ↗PDF ↗

摘要

Low-bit LLM quantization is useful only if a model preserves service-relevant accuracy, executes through supported GPU kernels, and meets measured memory and latency limits. We study Qwen3-8B on NVIDIA B200 with vLLM, including INT2/INT3, GPTQ, NVFP4, and mixed-precision configurations. We verify native packed execution and precision on all 144 fused linear modules and evaluate Korean, English, numerical, and reasoning quality under preregistered gates. Packed INT2 and INT3 execute, but all seven tested INT2 grids degenerate on generation. Isolated module-restoration effects are strongly non-additive; Korean and reasoning sensitivity rankings disagree, and a promising INT4/INT8 development candidate fails independent final evaluation. A configuration with 16-bit activations, 98 NVFP4 weight-only modules, 46 INT8 modules, and an INT8 output head passes all gates on the untouched FINAL-81 set (5,168 items): 5.917 GiB of weight-loading memory (limit 5.98), Korean teacher-correct retention 0.904 (gate 0.90), and TPOT ratio 0.939 relative to BF16. It improves paired accuracy over NVFP4 W4A4 by +1.4 percentage points (Bonferroni-adjusted interval [+0.2, +2.6]), compared with +3.4 points on the selection set. This comparison changes both activation precision and INT8 protection; exploratory analyses suggest that the activation change contributes substantially, but the isolated allocation effect is not established. Final checkpoints were not retained, and bitwise rebuild reproducibility remains unverified.

作者备注:18 pages, 2 figures

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

目前只收录了一个版本。收录下一版后,可以对照标题、作者、分类和摘要。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v1 · 2026-11-17 06:58 · 最新观测版本

Beyond Bitwidth: Deployment-Constrained Mixed-Precision Quantization for Multilingual LLM Inference on NVIDIA Blackwell

Low-bit LLM quantization is useful only if a model preserves service-relevant accuracy, executes through supported GPU kernels, and meets measured memory and latency limits. We study Qwen3-8B on NVIDIA B200 with vLLM, including INT2/INT3, GPTQ, NVFP4, and mixed-precision configurations. We verify native packed execution and precision on all 144 fused linear modules and evaluate Korean, English, numerical, and reasoning quality under preregistered gates. Packed INT2 and INT3 execute, but all seven tested INT2 grids degenerate on generation. Isolated module-restoration effects are strongly non-additive; Korean and reasoning sensitivity rankings disagree, and a promising INT4/INT8 development candidate fails independent final evaluation. A configuration with 16-bit activations, 98 NVFP4 weight-only modules, 46 INT8 modules, and an INT8 output head passes all gates on the untouched FINAL-81 set (5,168 items): 5.917 GiB of weight-loading memory (limit 5.98), Korean teacher-correct retention 0.904 (gate 0.90), and TPOT ratio 0.939 relative to BF16. It improves paired accuracy over NVFP4 W4A4 by +1.4 percentage points (Bonferroni-adjusted interval [+0.2, +2.6]), compared with +3.4 points on the selection set. This comparison changes both activation precision and INT8 protection; exploratory analyses suggest that the activation change contributes substantially, but the isolated allocation effect is not established. Final checkpoints were not retained, and bitwise rebuild reproducibility remains unverified.

查看此版本解读 →查看此版本原文 ↗