21 citations · 25 across the 4 of their papers we have counts for
5 papers
Eunoia: Toward Application-Transparent, Scalable GPU Sharing in AI-RAN
Zongshen Wu, Marko Milenković, Jiarong Xing +1
The AI-RAN paradigm makes the next-generation cellular network more versatile by co-locating AI and RAN workloads on the same commodity hardware. While the GPU is a promising platf…
EXP-Bench: Can AI Conduct AI Research Experiments?
Patrick Tser Jern Kon, Jiachen Liu, Xinyi Zhu +10
Automating AI research holds immense potential for accelerating scientific progress, yet current AI agents struggle with the complexities of rigorous, end-to-end experimentation. W…
Disaggregating Embedding Recommendation Systems with FlexEMR
Yibo Huang, Zhenning Yang, Jiarong Xing +5
Efficiently serving embedding-based recommendation (EMR) models remains a significant challenge due to their increasingly large memory requirements. Today's practice splits the mod…
Symbolic Distillation for Learned TCP Congestion Control
S P Sharan, Wenqing Zheng, Kuo-Feng Hsu +3
Recent advances in TCP congestion control (CC) have achieved tremendous success with deep reinforcement learning (RL) approaches, which use feedforward neural networks (NN) to lear…
Bolt: Bridging the Gap between Auto-tuners and Hardware-native Performance
Jiarong Xing, Leyuan Wang, Shang Zhang +3
Today's auto-tuners (e.g., AutoTVM, Ansor) generate efficient tensor programs by navigating a large search space to identify effective implementations, but they do so with opaque h…