# FluxMoE: Decoupling Expert Residency for High-Performance MoE Serving

arXiv: https://arxiv.org/abs/2604.02715 · HTML: https://mlsanity.com/paper/2604.02715/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Efficient MoE training and inference systems, one of 12 directions in Mixture-of-experts.
- Builds on 29 of its 44 references that are in the map.
- First posted on arXiv 3 Apr 2026.

## Placed in

- Mixture-of-experts / [Efficient MoE training and inference systems](https://mlsanity.com/moe/efficient-moe-training-and-inference-systems/)

## Builds on (most cited first, 29 of 29)

- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [Pointer Sentinel Mixture Models](https://mlsanity.com/paper/1609.07843/) (2016)
- [GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers](https://mlsanity.com/paper/2210.17323/) (2022)
- [AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration](https://mlsanity.com/paper/2306.00978/) (2023)
- [SGLang: Efficient Execution of Structured Language Model Programs](https://mlsanity.com/paper/2312.07104/) (2023)
- [DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Serving](https://mlsanity.com/paper/2401.09670/) (2024)
- [FlexGen: High-Throughput Generative Inference of Large Language Models with a Single GPU](https://mlsanity.com/paper/2303.06865/) (2023)
- [L1: Controlling How Long A Reasoning Model Thinks With Reinforcement Learning](https://mlsanity.com/paper/2503.04697/) (2025)
- [FlashInfer: Efficient and Customizable Attention Engine for LLM Inference Serving](https://mlsanity.com/paper/2501.01005/) (2025)
- [ZeRO-Infinity: Breaking the GPU Memory Wall for Extreme Scale Deep Learning](https://mlsanity.com/paper/2104.07857/) (2021)
- [CacheGen: KV Cache Compression and Streaming for Fast Large Language Model Serving](https://mlsanity.com/paper/2310.07240/) (2023)
- [Mooncake: A KVCache-centric Disaggregated Architecture for LLM Serving](https://mlsanity.com/paper/2407.00079/) (2024)
- [Pre-gated MoE: An Algorithm-System Co-Design for Fast and Scalable Mixture-of-Expert Inference](https://mlsanity.com/paper/2308.12066/) (2023)
- [DART-Math: Difficulty-Aware Rejection Tuning for Mathematical Problem-Solving](https://mlsanity.com/paper/2407.13690/) (2024)
- [NanoFlow: Towards Optimal Large Language Model Serving Throughput](https://mlsanity.com/paper/2408.12757/) (2024)
- [ChunkAttention: Efficient Self-Attention with Prefix-Aware KV Cache and Two-Phase Partition](https://mlsanity.com/paper/2402.15220/) (2024)
- [MoE-Lightning: High-Throughput MoE Inference on Memory-constrained GPUs](https://mlsanity.com/paper/2411.11217/) (2024)
- [MoE-Infinity: Efficient MoE Inference on Personal Machines with Sparsity-Aware Expert Cache](https://mlsanity.com/paper/2401.14361/) (2024)
- [Taming Latency-Memory Trade-Off in MoE-Based LLM Serving via Fine-Grained Expert Offloading](https://mlsanity.com/paper/2502.05370/) (2025)
- [70% Size, 100% Accuracy: Lossless LLM Compression for Efficient GPU Inference via Dynamic-Length Float (DFloat11)](https://mlsanity.com/paper/2504.11651/) (2025)
- [Klotski: Efficient Mixture-of-Expert Inference via Expert-Aware Multi-Batch Pipeline](https://mlsanity.com/paper/2502.06888/) (2025)
- [ZipNN: Lossless Compression for AI Models](https://mlsanity.com/paper/2411.05239/) (2024)
- [Fate: Fast Edge Inference of Mixture-of-Experts Models via Cross-Layer Gate](https://mlsanity.com/paper/2502.12224/) (2025)
- [eMoE: Task-aware Memory Efficient Mixture-of-Experts-Based (MoE) Model Inference](https://mlsanity.com/paper/2503.06823/) (2025)
- [ZipServ: Fast and Memory-Efficient LLM Inference with Hardware-Aware Lossless Compression](https://mlsanity.com/paper/2603.17435/) (2026)
- [ZipMoE: Efficient On-Device MoE Serving via Lossless Compression and Cache-Affinity Scheduling](https://mlsanity.com/paper/2601.21198/) (2026)
- [Lossless Compression of Neural Network Components: Weights, Checkpoints, and K/V Caches in Low-Precision Formats](https://mlsanity.com/paper/2508.19263/) (2025)
- [MELINOE: Fine-Tuning Enables Memory-Efficient Inference for Mixture-of-Experts Models](https://mlsanity.com/paper/2602.11192/) (2026)

## Abstract (arXiv)

Mixture-of-Experts (MoE) models have become mainstream for scaling language models to hundreds of billions of expert parameters. Despite sparse expert activation, existing inference engines keep all experts GPU-resident, crowding out the key-value cache in large-batch, long-output offline workloads. We present FluxMoE, which decouples experts from physical GPU residency and adapts their footprint to available memory through a new expert paging abstraction. FluxMoE combines PagedTensor for transparent remapping, a bandwidth-balanced hierarchy spanning losslessly compressed GPU memory and host DRAM, and a budget-aware residency planner. Unlike CPU-GPU co-inference and whole-layer offloading, FluxMoE streams weights on demand while keeping expert computation on GPUs. We implement FluxMoE atop vLLM and evaluate it on three MoE models. For GLM-4.5 on 8×H20 GPUs, FluxMoE delivers up to 7.2× vLLM's throughput and 79.0% lower average Time-Per-Output-Token (TPOT), without measurable model-quality loss using lossless compression. For Mixtral-8×7B-Instruct on 2×L40S GPUs, where weight-resident vLLM cannot fit, FluxMoE delivers 4.3× KTransformers's throughput and 29.1% lower average TPOT.