Computing in the Era of Large Generative Models: From Cloud-Native to AI-Native

no code implementations • 17 Jan 2024 • Yao Lu, Song Bian, Lequn Chen, Yongjun He, Yulong Hui, Matthew Lentz, Beibin Li, Fei Liu, Jialin Li, Qi Liu, Rui Liu, Xiaoxuan Liu, Lin Ma, Kexin Rong, Jianguo Wang, Yingjun Wu, Yongji Wu, Huanchen Zhang, Minjia Zhang, Qizhen Zhang, Tianyi Zhou, Danyang Zhuo

In this paper, we investigate the intersection of large generative AI models and cloud-native computing architectures.

SRL: Scaling Distributed Reinforcement Learning to Over Ten Thousand Cores

1 code implementation • 29 Jun 2023 • Zhiyu Mei, Wei Fu, Guangju Wang, Huanchen Zhang, Yi Wu

In a large-scale cluster, the novel architecture of SRL leads to up to 3. 7x speedup compared to the design choices adopted by the existing libraries.

reinforcement-learning Reinforcement Learning (RL)

LeCo: Lightweight Compression via Learning Serial Correlations

no code implementations • 27 Jun 2023 • Yihao Liu, Xinyu Zeng, Huanchen Zhang

Lightweight data compression is a key technique that allows column stores to exhibit superior performance for analytical queries.

Data Compression

Proteus: A Self-Designing Range Filter

2 code implementations • 30 Jun 2022 • Eric R. Knorr, Baptiste Lemaire, Andrew Lim, Siqiang Luo, Huanchen Zhang, Stratos Idreos, Michael Mitzenmacher

We introduce Proteus, a novel self-designing approximate range filter, which configures itself based on sampled data in order to optimize its false positive rate (FPR) for a given space requirement.