*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Hyesoon Kim
Assistant Professor, School of Computer Science, Georgia Tech
For more information please contact Dr. Rich Vuduc at richie@cc.gatech.edu
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
"An analytical GPU performance model and a dynamic compilation system for CPU/GPU systems"
Abstract:
Heterogeneous multiprocessors are increasingly important in the multi-core era due to their potential for high performance and energy efficiency. To increase efficiency, it is very critical to find optimal resources for different computations. In this talk, I’ll present two our on-going work about building an efficient heterogeneous computing system.
First, I’ll present a static time GPU performance analytical model. The GPU analytical model provides insights into the performance bottlenecks of parallel applications on GPU architectures. The key component of the model is estimating the number of parallel memory requests by considering the number of running threads and memory bandwidth. Through this analytical model, we can predict the execution time of a set of important GPGPU applications on average less than a 13% error.
Secondly, I’ll present an automatic technique to map computations to processing elements on a CPU+GPU machine, which we call Qilin. Qilin dynamically compiles code and distributes the computation between the CPU and GPU automatically. This automatic adaptive mapping achieves a 25% reduction in execution time and a 20% reduction in energy consumption than static mappings for a set of important GPGPU applications.
Bio:
Hyesoon Kim is an Assistant professor in the School of Computer Science at the Georgia Institute of Technology. Her research interests include high-performance energy-efficient heterogeneous architectures, programmer-compiler-microarchitecture interaction especially for CPU/GPU systems. She received a BA in mechanical engineering from Korea Advanced Institute of Science and Technology (KAIST), an MS in mechanical engineering from Seoul National University, and an MS and a Ph.D in computer engineering at The University of Texas at Austin.
~~~~~~~~~~~
You are cordially invited to attend a reception between 1:30 - 2:00; before the seminar to chat informally with faculty and students.
Refreshments will be provided
To receive future announcements, please sign up to the cse-seminar email list:
https://mailman.cc.gatech.edu/mailman/listinfo/cse-seminar
