4 papers
KVServe: Service-Aware KV Cache Compression for Communication-Efficient Disaggregated LLM Serving
Zedong Liu, Xinyang Ma, Dejun Luo +9
LLMs are widely adopted in production, pushing inference systems to their limits. Disaggregated LLM serving (e.g., PD separation and KV state disaggregation) improves scalability a…
CCL-D: A High-Precision Diagnostic System for Slow and Hang Anomalies in Large-Scale Model Training
Yida Gu, Fakang Wang, Jianhao Fu +17
As training scales grow, collective communication libraries (CCL) increasingly face anomalies arising from complex interactions among hardware, software, and environmental factors.…
TACO: Efficient Communication Compression of Intermediate Tensors for Scalable Tensor-Parallel LLM Training
Man Liu, Xingchen Liu, Xingjian Tian +8
Handling communication overhead in large-scale tensor-parallel training remains a critical challenge due to the dense, near-zero distributions of intermediate tensors, which exacer…
ENEC: A Lossless AI Model Compression Method Enabling Fast Inference on Ascend NPUs
Jinwu Yang, Jiaan Wu, Zedong Liu +17
The rapid scaling of Large Language Models presents significant challenges for their deployment and inference, particularly on resource-constrained specialized AI hardware accelera…