D
Daniel Crankshaw
Researcher at University of California, Berkeley
Publications - 16
Citations - 1999
Daniel Crankshaw is an academic researcher from University of California, Berkeley. The author has contributed to research in topics: Robustness (computer science) & Hardware acceleration. The author has an hindex of 10, co-authored 14 publications receiving 1686 citations. Previous affiliations of Daniel Crankshaw include Johns Hopkins University & Microsoft.
Papers
More filters
Proceedings ArticleDOI
GraphX: graph processing in a distributed dataflow framework
TL;DR: This paper introduces GraphX, an embedded graph processing framework built on top of Apache Spark, a widely used distributed dataflow system and demonstrates that GraphX achieves an order of magnitude performance gain over the base dataflow framework and matches the performance of specialized graph processing systems while enabling a wider range of computation.
Proceedings Article
Clipper: a low-latency online prediction serving system
TL;DR: Clipper is introduced, a general-purpose low-latency prediction serving system that introduces a modular architecture to simplify model deployment across frameworks and applications and improves prediction throughput, accuracy, and robustness without modifying the underlying machine learning frameworks.
Posted Content
Clipper: A Low-Latency Online Prediction Serving System
TL;DR: Clipper as mentioned in this paper is a general-purpose low-latency prediction serving system, which is able to meet the latency, accuracy, and throughput demands of online serving applications by introducing caching, batching, and adaptive model selection techniques.
Proceedings Article
The Missing Piece in Complex Analytics: Low Latency, Scalable Model Management and Serving with Velox
Daniel Crankshaw,Peter Bailis,Joseph E. Gonzalez,Haoyuan Li,Zhao Zhang,Michael J. Franklin,Ali Ghodsi,Michael I. Jordan +7 more
TL;DR: The challenges and architectural considerations required to achieve this functionality, including the abilities to span online and offline systems, to adaptively adjust model materialization strategies, and to exploit inherent statistical properties such as model error tolerance, all while operating at "Big Data" scale are described.
Posted Content
GraphX: Unifying Data-Parallel and Graph-Parallel Analytics
TL;DR: It is demonstrated that GraphX achieves comparable performance as specialized graph computation systems, while outperforming them in end-to-end graph pipelines, achieving a balance between expressiveness, performance, and ease of use.