Copertina

FlashMoE: Reducing SSD I/O Bottlenecks via ML-Based Cache Replacement for Mixture-of-Experts Inference on Edge Devices

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Kim, Byeongju, Lee, Jungwan, Han, Donghyeon, Yoo, Hoi-Jun, Kim, Sangyeob
Natura:	Preprint
Pubblicazione:	2026
Soggetti:	Machine Learning Artificial Intelligence
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

SliceMoE: Bit-Sliced Expert Caching under Miss-Rate Constraints for Efficient MoE Inference
di: Choi, Yuseon, et al.
Pubblicazione: (2025)

FlashMoE: Fast Distributed MoE in a Single Kernel
di: Aimuyo, Osayamen Jonathan, et al.
Pubblicazione: (2025)

LightMoE: Reducing Mixture-of-Experts Redundancy through Expert Replacing
di: Hao, Jiawei, et al.
Pubblicazione: (2026)

SlimCaching: Edge Caching of Mixture-of-Experts for Distributed Inference
di: Chen, Qian, et al.
Pubblicazione: (2025)

Rethinking LLM Inference Bottlenecks: Insights from Latent Attention and Mixture-of-Experts
di: Yun, Sungmin, et al.
Pubblicazione: (2025)

Mixture of Cache-Conditional Experts for Efficient Mobile Device Inference
di: Skliar, Andrii, et al.
Pubblicazione: (2024)

Flash Communication: Reducing Tensor Parallelization Bottleneck for Fast Large Language Model Inference
di: Li, Qingyuan, et al.
Pubblicazione: (2024)

MoFE: Mixture of Frozen Experts Architecture
di: Seo, Jean, et al.
Pubblicazione: (2025)

Mixture of Concept Bottleneck Experts
di: De Santis, Francesco, et al.
Pubblicazione: (2026)

MoBiLE: Efficient Mixture-of-Experts Inference on Consumer GPU with Mixture of Big Little Experts
di: Zhao, Yushu, et al.
Pubblicazione: (2025)

MobileMoE: Scaling On-Device Mixture of Experts
di: Chen, Yanbei, et al.
Pubblicazione: (2026)

Flash Cache: Reducing Bias in Radiance Cache Based Inverse Rendering
di: Attal, Benjamin, et al.
Pubblicazione: (2024)

Rethinking LiDAR Domain Generalization: Single Source as Multiple Density Domains
di: Kim, Jaeyeul, et al.
Pubblicazione: (2023)

ExpertFlow: Efficient Mixture-of-Experts Inference via Predictive Expert Caching and Token Scheduling
di: He, Xin, et al.
Pubblicazione: (2024)

Faster LLM Inference using DBMS-Inspired Preemption and Cache Replacement Policies
di: Kim, Kyoungmin, et al.
Pubblicazione: (2024)

SeVeDo: A Heterogeneous Transformer Accelerator for Low-Bit Inference via Hierarchical Group Quantization and SVD-Guided Mixed Precision
di: Choi, Yuseon, et al.
Pubblicazione: (2025)

SeaCache: Spectral-Evolution-Aware Cache for Accelerating Diffusion Models
di: Chung, Jiwoo, et al.
Pubblicazione: (2026)

ELMoE-3D: Leveraging Intrinsic Elasticity of MoE for Hybrid-Bonding-Enabled Self-Speculative Decoding in On-Premises Serving
di: Choi, Yuseon, et al.
Pubblicazione: (2026)

SSD Offloading for LLM Mixture-of-Experts Weights Considered Harmful in Energy Efficiency
di: Kyung, Kwanhee, et al.
Pubblicazione: (2025)

ReinPool: Reinforcement Learning Pooling Multi-Vector Embeddings for Retrieval System
di: Cha, Sungguk, et al.
Pubblicazione: (2026)

An Empirical Study of Group Conformity in Multi-Agent Systems
di: Choi, Min, et al.
Pubblicazione: (2025)

On the normalized local volume of a non-closed point
di: Kim, Donghyeon
Pubblicazione: (2026)

Asymptotically flat divisors and strongly $F$-regular type varieties
di: Kim, Donghyeon
Pubblicazione: (2025)

On a cohomological property of the center of a resolution
di: Kim, Donghyeon
Pubblicazione: (2022)

The number of smooth varieties in an MMP on a 3-fold of Fano type
di: Kim, Donghyeon
Pubblicazione: (2025)

On diminished multiplier ideal and the termination of flips
di: Kim, Donghyeon
Pubblicazione: (2024)

On volumes and the generic invariance of Fano type varieties
di: Kim, Donghyeon
Pubblicazione: (2025)

Rational singularities and $q$-birational morphism
di: Kim, Donghyeon
Pubblicazione: (2023)

CoX-MoE: Coalesced Expert Execution for High-Throughput MoE Inference with AMX-Enabled CPU-GPU Co-Execution
di: Son, Muyoung, et al.
Pubblicazione: (2026)

PreMoE: Proactive Inference for Efficient Mixture-of-Experts
di: Pei, Zehua, et al.
Pubblicazione: (2025)

Neuromorphic Wireless Device-Edge Co-Inference via the Directed Information Bottleneck
di: Ke, Yuzhen, et al.
Pubblicazione: (2024)

Mixture of Mini Experts: Overcoming the Linear Layer Bottleneck in Multiple Instance Learning
di: Shao, Daniel, et al.
Pubblicazione: (2026)

A 0.5V, 6.2$μ$W, 0.059mm$^{2}$ Sinusoidal Current Generator IC with 0.088% THD for Bio-Impedance Sensing
di: Kim, Kwantae, et al.
Pubblicazione: (2024)

Scaling Multi-Node Mixture-of-Experts Inference Using Expert Activation Patterns
di: Bambhaniya, Abhimanyu, et al.
Pubblicazione: (2026)

MoNDE: Mixture of Near-Data Experts for Large-Scale Sparse Models
di: Kim, Taehyun, et al.
Pubblicazione: (2024)

Mustafar: Promoting Unstructured Sparsity for KV Cache Pruning in LLM Inference
di: Joo, Donghyeon, et al.
Pubblicazione: (2025)

MoE-GS: Mixture of Experts for Dynamic Gaussian Splatting
di: Jin, In-Hwan, et al.
Pubblicazione: (2025)

ZipCache: A DRAM/SSD Cache with Built-in Transparent Compression
di: Xie, Rui, et al.
Pubblicazione: (2024)

MoE-GRPO: Optimizing Mixture-of-Experts via Reinforcement Learning in Vision-Language Models
di: Ko, Dohwan, et al.
Pubblicazione: (2026)

MoDEM: Mixture of Domain Expert Models
di: Simonds, Toby, et al.
Pubblicazione: (2024)