Dynamic Expert Quantization for Scalable Mixture-of-Experts Inference

• URL: http://arxiv.org/abs/2511.15015v2
• Date: Mon, 24 Nov 2025 00:36:49 GMT
• Title: Dynamic Expert Quantization for Scalable Mixture-of-Experts Inference
• Authors: Kexin Chu, Dawei Xiang, Zixu Shen, Yiwei Yang, Zecheng Liu, Wei Zhang,
• Abstract summary: We present DynaExq, a runtime system that treats expert precision as a first-class, dynamically managed resource.<n>We show that DynaExq deploys large LLMs on single 5090 and A6000 GPUs and improves accuracy by up to 4.03 points over static low-precision baselines.
• Score: 2.649774320778185
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Mixture-of-Experts (MoE) models scale LLM capacity efficiently, but deployment on consumer GPUs is limited by the large memory footprint of inactive experts. Static post-training quantization reduces storage costs but cannot adapt to shifting activation patterns, causing accuracy loss under aggressive compression. So we present DynaExq, a runtime system that treats expert precision as a first-class, dynamically managed resource. DynaExq combines (1) a hotness-aware precision controller that continuously aligns expert bit-widths with long-term activation statistics, (2) a fully asynchronous precision-switching pipeline that overlaps promotion and demotion with MoE computation, and (3) a fragmentation-free memory pooling mechanism that supports hybrid-precision experts with deterministic allocation. Together, these components enable stable, non-blocking precision transitions under strict HBM budgets. Across Qwen3-30B and Qwen3-80B MoE models and six representative benchmarks, DynaExq deploys large LLMs on single RTX 5090 and A6000 GPUs and improves accuracy by up to 4.03 points over static low-precision baselines. The results show that adaptive, workload-aware quantization is an effective strategy for memory-constrained MoE serving.

Related papers

• Advancing Model Refinement: Muon-Optimized Distillation and Quantization for LLM Deployment [1.0858059444801136]
  Large Language Models (LLMs) enable advanced natural language processing but face deployment challenges on resource-constrained edge devices.<n>We propose an integrated framework combining GPTQ-based quantization, low-rank adaptation (LoRA), and a specialized data distillation process.
  arXiv  Detail & Related papers  (2026-01-14T20:50:30Z)
• MC#: Mixture Compressor for Mixture-of-Experts Large Models [86.64315380917827]
  Mixture-of-Experts (MoE) effectively scales large language models (LLMs) and vision-language models (VLMs) by increasing capacity through sparse activation.<n>We propose MC# (Mixture-Compressor-sharp), a framework that combines static quantization and dynamic expert pruning.
  arXiv  Detail & Related papers  (2025-10-13T03:12:46Z)
• The Curious Case of In-Training Compression of State Space Models [49.819321766705514]
  State Space Models (SSMs) tackle long sequence modeling tasks efficiently, offer both parallelizable training and fast inference.<n>Key design challenge is striking the right balance between maximizing expressivity and limiting this computational burden.<n>Our approach, textscCompreSSM, applies to Linear Time-Invariant SSMs such as Linear Recurrent Units, but is also extendable to selective models.
  arXiv  Detail & Related papers  (2025-10-03T09:02:33Z)
• PT$^2$-LLM: Post-Training Ternarization for Large Language Models [52.4629647715623]
  Large Language Models (LLMs) have shown impressive capabilities across diverse tasks, but their large memory and compute demands hinder deployment.<n>We propose PT$2$-LLM, a post-training ternarization framework tailored for LLMs.<n>At its core is an Asymmetric Ternary Quantizer equipped with a two-stage refinement pipeline.
  arXiv  Detail & Related papers  (2025-09-27T03:01:48Z)
• FlexQuant: A Flexible and Efficient Dynamic Precision Switching Framework for LLM Quantization [19.12288373558071]
  We propose FlexQuant, a dynamic precision-switching framework to optimize the trade-off between inference speed and accuracy.<n>We show that FlexQuant achieves a 1.3x end-to-end speedup across diverse language tasks with negligible accuracy loss.
  arXiv  Detail & Related papers  (2025-05-21T07:42:53Z)
• Quaff: Quantized Parameter-Efficient Fine-Tuning under Outlier Spatial Stability Hypothesis [9.884521812433661]
  Quaff is a Quantized parameter-efficient fine-tuning framework for large language models.<n>It suppresses outliers exclusively invariant channels using lightweight operations.<n>It achieves a 1.73x latency reduction and 30% memory savings over full-precision fine-tuning.
  arXiv  Detail & Related papers  (2025-05-20T07:19:36Z)
• D$^{2}$MoE: Dual Routing and Dynamic Scheduling for Efficient On-Device MoE-based LLM Serving [14.607254882119507]
  Combination of experts (MoE) model is a sparse variant of large language models (LLMs)<n>Despite its benefits, MoE is still too expensive to deploy on resource-constrained edge devices.<n>We propose D$2$MoE, an algorithm-system co-design framework that matches diverse task requirements by dynamically allocating the most proper bit-width to each expert.
  arXiv  Detail & Related papers  (2025-04-17T05:37:35Z)
• HOBBIT: A Mixed Precision Expert Offloading System for Fast MoE Inference [54.40808356999408]
  We present HOBBIT, a mixed precision expert offloading system to enable flexible and efficient MoE inference. Our key insight is that dynamically replacing less critical cache-miss experts with low precision versions can substantially reduce expert-loading latency. HOBBIT achieves up to a 9.93x speedup in decoding compared to state-of-the-art MoE offloading systems.
  arXiv  Detail & Related papers  (2024-11-03T04:25:46Z)
• SliM-LLM: Salience-Driven Mixed-Precision Quantization for Large Language Models [63.118592279833656]
  Post-training quantization (PTQ) is an effective technique for compressing large language models (LLMs)<n>We propose SliM-LLM, a salience-driven mixed-precision quantization framework that allocates bit-widths at the group-wise.<n> Experiments show that SliM-LLM achieves superior performance across various LLMs at low bit-widths.
  arXiv  Detail & Related papers  (2024-05-23T16:21:48Z)
• SqueezeLLM: Dense-and-Sparse Quantization [80.32162537942138]
  Main bottleneck for generative inference with LLMs is memory bandwidth, rather than compute, for single batch inference. We introduce SqueezeLLM, a post-training quantization framework that enables lossless compression to ultra-low precisions of up to 3-bit. Our framework incorporates two novel ideas: (i) sensitivity-based non-uniform quantization, which searches for the optimal bit precision assignment based on second-order information; and (ii) the Dense-and-Sparse decomposition that stores outliers and sensitive weight values in an efficient sparse format.
  arXiv  Detail & Related papers  (2023-06-13T08:57:54Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.