NanoCP introduces request-level dynamic context parallelism to decouple MoE communication from KV cache placement in hybrid data-expert parallel serving, reporting up to 3.27x higher request rates and 2.12x lower P99 latency under TPOT SLOs.
arXiv:2503.19779 [cs.LG]https://arxiv.org/abs/2503.19779
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
representative citing papers
A hybrid JIT-CUDA Graph framework reduces TTFT by up to 66% and P99 latency versus TensorRT-LLM for single-GPU LLaMA-2 7B inference on short prompts.
citing papers explorer
-
NanoCP: Request-Level Dynamic Context Parallelism for Data-Expert Parallel Decoding
NanoCP introduces request-level dynamic context parallelism to decouple MoE communication from KV cache placement in hybrid data-expert parallel serving, reporting up to 3.27x higher request rates and 2.12x lower P99 latency under TPOT SLOs.
-
Hybrid JIT-CUDA Graph Optimization for Low-Latency Large Language Model Inference
A hybrid JIT-CUDA Graph framework reduces TTFT by up to 66% and P99 latency versus TensorRT-LLM for single-GPU LLaMA-2 7B inference on short prompts.