Cymbalo: An Efficient Graph Processing Framework for Machine Learning
| Autor: | Biwei Xie, Xinhui Tian, Jianfeng Zhan |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
020203 distributed computing
Computer science business.industry Computation 02 engineering and technology Machine learning computer.software_genre Graph 020204 information systems 0202 electrical engineering electronic engineering information engineering Programming paradigm Memory footprint Graph (abstract data type) Artificial intelligence business computer |
| Zdroj: | ISPA/IUCC/BDCloud/SocialCom/SustainCom |
| DOI: | 10.1109/bdcloud.2018.00090 |
| Popis: | Due to the growth of data scale, distributed machine learning has become more important than ever. Some recent work, like TuX^2, show promising prospect in dealing with distributed machine learning by leveraging the power of graph computation, but still leave some key problems unsolved. In this paper, we propose Cymbalo, a new distributed graph processing framework for large-scale machine learning algorithms. To satisfy the specific characteristics of machine learning, Cymbalo employs a heterogeneity-aware data model, a hybrid computing model and a vector-aware programming model, to ensure small memory footprint, good computation efficiency and expressiveness. The experiment results show that Cymbalo outperforms Spark by 2.4×-3.2×, and PowerGraph by up to 5.8×. Moreover, Cymbalo can also outperform Angel, a recent parameter server system, by 1.6×-2.1×. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|