Monte Carlo simulation benchmark for MoE expert activation memory variability and admission control under non-deterministic routing — including routing skew amplification, top_k sweep, and policy comparison across uniform, skewed, and very_skewed distributions
performance-engineering benchmark simulation scheduling inference systems moe monte-carlo-simulation mixture-of-experts admission-control llm llm-serving machine-learning-infrastructure expert-routing serving-systems load-imbalance activation-memory routing-skew
-
Updated
Jul 26, 2026 - Python