paper-with-me

홈 › Papers

Rethinking Network Topologies for Cost-Effective Mixture-of-Experts LLM Serving

2026-04-30 · Junsun Choi, Sam Son, Sunjin Choi, Hansung Kim, Yakun Sophia Shao, Scott Shenker, Sylvia Ratnasamy, Borivoje Nikolic arxiv

Mixture-of-experts (MoE) architectures have turned LLM serving into a cluster-scale workload in which communication consumes a considerable portion of LLM serving runtime. This has prompted industry to invest heavily in expensive high-bandwidth scale-up networks. We question whether such costly infrastructure is strictly necessary. We present the first systematic cross-layer analysis of network cost-effectiveness for MoE LLM serving, comparing four representative XPU (e.g., GPU/TPU) topologies (scale-up, scale-out, 3D torus, and 3D full-mesh). We find that lower-cost switchless topologies are more cost-effective than the scale-up topology across all serving scenarios explored, improving cost-effectiveness by 20.6-56.2%. In particular, the 3D full-mesh topology is Pareto-optimal in terms of the performance-cost tradeoff. We also find that current scale-up link bandwidths are over-provisioned: reducing the link bandwidth improves throughput per cost by up to 27%. A forward-looking analysis of upcoming GPU generations indicates that the cost-performance advantage of switchless networks will likely persist.

📄 PDF Abstract BibTeX arXiv:2605.00254

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

MoE-ViE: Mixture of Experts Vision Encoder for Efficient Image and Video Understanding

2026-08-18 · Bonan Zhang, Shiyu Dong, Quan Hung Tran, Katharina Gschwind 외 arxiv

Vision encoders are a critical component of vision-language models, and scaling their capacity effectively improves performance. However, dense scaling increases compute cost and inference latency. Mixture-of-Experts (Mo…

Scaling Intelligence: Designing Data Centers for Next-Gen Language Models

2025-06-17 · Jesmin Jahan Tithi, Hanjiang Wu, Avishaii Abuhatzera, Fabrizio Petrini

The explosive growth of Large Language Models (LLMs) - such as GPT-4 with 1.8 trillion parameters - demands a radical rethinking of data center architecture to ensure scalability, efficiency, and cost-effectiveness. Our …

Mixture-of-Experts

AxMoE: Characterizing the Impact of Approximate Multipliers on Mixture-of-Experts DNN Architectures

2026-05-06 · Omkar B Shende, Marcello Traiola, Gayathri Ananthanarayanan arxiv

Deep neural network (DNN) inference at the edge demands simultaneous improvements in accuracy, computational efficiency, and energy consumption. Approximate computing and Mixture-of-Experts (MoE) architectures have each …

Computational Efficiency

GRAPHMOE: Amplifying Cognitive Depth of Mixture-of-Experts Network via Introducing Self-Rethinking Mechanism

2025-01-14 · Chen Tang, Bo Lv, Zifan Zheng, Bohao Yang 외

Traditional Mixture-of-Experts (MoE) networks benefit from utilizing multiple smaller expert models as opposed to a single large network. However, these experts typically operate independently, leaving a question open ab…

Mixture-of-Experts

MoNTA: Accelerating Mixture-of-Experts Training with Network-Traffc-Aware Parallel Optimization

2024-11-01 · Jingming Guo, Yan Liu, Yu Meng, Zhiwei Tao 외

The Mixture of Experts (MoE) is an advanced model architecture in the industry that combines multiple specialized expert models from various domains into a single supermodel. This approach enables the model to scale with…

8kMixture-of-Experts