SmartApe AI
Oct 11, 2026 · Artificial Intelligence
AInfer-PD: In-Place Prefill-Decode Multiplexing for Distributed MoE Agentic RL Rollout
Ant Group's AInfer-PD enables in-place Prefill-Decode multiplexing for distributed MoE models in Agentic RL rollouts, solving cross-rank collective communication ordering with Rank-Aligned Segment Turnstile and DeepEP state conflicts by separating mutable communication states, reducing rollout completion time by up to 32.9% versus SGLang on H20 GPUs.
AInfer-PDCollective CommunicationDeepEP
0 likes · 10 min read
