[2017-12-01] Dr.Jilong Xue, Microsoft, “Optimizing Deep Learning Computation over Modern Hardware”


Title: Optimizing Deep Learning Computation over Modern Hardware
Date: 2017-12-01  02:20pm-03:30pm
Location: R103, CSIE
Speaker:  Dr.Jilong Xue, Microsoft
Hosted by: Prof. Winston Hsu


Deep learning emerges as an important new resource-intensive workload and has been successfully applied in computer vision, speech, natural language processing, and so on. Large-scale deep learning computation for model training is becoming a necessity to cope with the ever-growing data and model sizes. Deep learning computation is typically characterized by a simple tensor data abstraction to model multidimensional matrices, a data-flow graph to model computation, and iterative executions with relatively frequent synchronizations, thereby making it substantially different from Map/Reduce style distributed big data computation.


This talk will first discuss the evolvement of the emerging deep learning frameworks and design choices, then introduce our recent research effort on Wolong, an optimization framework for deep learning computation. Wolong targets to build a deep learning backend that can provide automatic optimization for both scalability and efficiency of distributed and local execution. It applies RDMA-aware data-flow graph analysis to optimize the distributed execution plan to achieve the efficient communication. It also conducts automatic operator batching and kernel fusion to avoid operator scheduling overhead in local execution on GPU. Currently, Wolong can transparently improve deep learning computation performance by up to 8 times.



Jilong Xue is a researcher in System Research Group of Microsoft Research Asia (MSRA). His research focuses on building large-scale computing systems for resource-intensive workloads such as machine learning, deep learning, etc. through leveraging modern hardware including RDMA, GPU and ASIC. Recently, he is actively working in the areas of optimizing deep learning framework to bridge AI applications and diverse hardware resources, as well as building the next generation systems for future AI workloads. Jilong received his Ph.D. in Computer Science from Peking University in 2016. He had been working on large-scale graph processing system, social network security, and streaming system during his PhD study.



 (short update about MSRA),

Title: A glimpse of Microsoft Research Asia
Speaker:  Dr. Winnie Cui Microsoft



Microsoft Research Asia is Microsoft’s fundamental research arm in the Asia Pacific region. Today, with more than 200 researchers and developers and more than 300 visiting scientists and students, the lab conducts basic and applied research in areas central to Microsoft’s long-term strategy and future computing vision. It also collaborates with many universities in Asia. How does it collaborate? Does it have internship program? Who qualifies for its Fellowship award? I’ll tell you more in this quick update about MSRA.



Winnie Cui is a senior manager in the Academic Outreach organization at Microsoft Research Asia (MSRA), responsible for promoting and championing academic collaborations and partnership between MSRA and top-tier universities in Hong Kong, Taiwan and Singapore. Since joining Microsoft in 2006, Winnie has served in various positions, including Principal Solution Manager and Area Manager for North Asia in Microsoft IT. Winnie co-authored the chapter “Fluid Dynamics and Thrombosis” in the book “Advances in Cardiovascular Engineering” published by Springer. Winnie conducted her postdoctoral research at the University of Southern California and holds MS and PhD degrees in biomedical engineering from the University of Tennessee in the United States.




最後修改時間:2017-11-24 PM 4:39

cron web_use_log