Menu
View
Theme
Text size
The paper in your inbox every morning
Straight from the source · Glossary

The Tech Gazette

Mixture of experts (MoE)

Vol. I No. 26 Saturday, October 17, 2026 Closing 12:53 PM UTC · next edition 10:00 AM UTC

A model split into many sub-networks (experts), of which only a few run for each token, so it can be large but cheaper to run.

Read more at en.wikipedia.org →

Learn it

  1. NVIDIA DeveloperOct 15, 2026Technical Guides
    Scale Agentic Coding Workflows with Meta’s Muse Spark Open on NVIDIA GB300 NVL72
  2. AWS Machine LearningOct 13, 2026Technical Guides
    Mixture-of-Experts kernel engineering on Amazon SageMaker AI
  3. AWS Machine LearningSep 25, 2026Technical Guides
    Scaling MoE reinforcement learning on Amazon EKS with EFA and DeepEP with 40% more throughput
  4. NVIDIA DeveloperSep 24, 2026Technical Guides
    Efficient MoE Training for Biological Foundation Models
  5. AMD ROCmSep 22, 2026Technical Guides
    Benchmarking Kimi-K3 Across vLLM, SGLang, and ATOM on MI350X

In the news

  1. llama.cppOct 17, 2026Releases
    b11709
  2. OpenRouterOct 17, 2026New Models
    New model: DeepSeek: DeepSeek V4.1 Flash 1017
  3. OpenRouterOct 16, 2026New Models
    New model: Reflection: Beam 1.0 501B A23B
  4. BasetenOct 16, 2026News
    Beam: Efficient reasoning for coding and agentic workloads
  5. OpenRouterOct 15, 2026New Models
    New model: Moondream: Moondream 3 Preview
  6. OpenRouterOct 15, 2026New Models
    New model: Moondream: Moondream 3.1
  7. OllamaOct 15, 2026Releases
    Ollama 0.40.5: NVIDIA Nemotron 3.5 Super
  8. TransformersOct 15, 2026Releases
    Release v5.20.0
  9. arXiv cs.LGOct 15, 2026Research
    Rethinking Routing Exploration in Modern Mixture-of-Experts
  10. BasetenOct 15, 2026News
    Announcing NVIDIA Nemotron 3.5 Super VL on Baseten
  11. OpenRouterOct 15, 2026New Models
    New model: NVIDIA: Nemotron 3.5 Super VL
  12. Hugging FaceOct 15, 2026News
    Welcome Muse Spark Open: open-weight frontier multimodal mixture-of-experts model from Meta
  13. vLLMOct 15, 2026News
    Day 0 support for Muse Spark Open on vLLM
  14. arXiv cs.LGOct 14, 2026Research
    Beyond Expert Rankings: Contrast Calibration for Mixed-Precision MoE Quantization
  15. UnslothOct 13, 2026Releases
    Large Performance Gains + More Browser Use
  16. OpenRouterOct 13, 2026New Models
    New model: LiquidAI: LFM2.5-24B-A2B
  17. arXiv (systems & security) cs.DCOct 13, 2026Research
    FastEEP: Fast Intra-Instance Elastic Scaling for Expert-Parallel MoE Serving
  18. OpenRouterOct 12, 2026New Models
    New model: Nace.AI: Drex v2.0
  19. PyTorch · YouTubeOct 10, 2026Watch
    MoRI + vLLM: Wide Expert Parallelism and RDMA KV-Cache Transfer for Disaggregated MoE Serving on AMD
  20. OpenRouterOct 9, 2026New Models
    New model: Cloudflare: Clef Omni