dynamic operator scheduling 1heterogeneous platforms 1large language model inference 1neural processing units 1processing-in-memory 1weight layout optimization 1
From the 1 of 2 linked papers with an AI index.
2 papers
cs.AR2026
MCHA: A Memory-Centric Hierarchical Architecture for Parallel-Sequential Computing
Daijing Shi, Hongxiao Zhao, Yihan Fu +7
Emerging workloads, such as Multi-Agent Reinforcement Learning (MARL), large-scale neuromorphic computing, and probabilistic graphical models, intrinsically exhibit parallel-sequen…
cs.AR2026
Beyond Prefill-Decode Disaggregation: Dissecting LLM Inference for Heterogeneous Platforms via Dynamic Operator Scheduling
Jiaqi Yang, Jiayi Li, Yihan Fu +5
The paper introduces DOPS, a framework that dynamically schedules LLM operators and chooses efficient weight layouts to improve inference latency on heterogeneous systems with NPUs…