Seminar Schedule

Seminar Home   Hide Abstracts   PDF   Plain Text
Chronological By Speaker
 
Title Sequential Detection of Network Intrusions
Speaker Dr. Xinjia Chen
Dept. of Electrical Engineering, Southern University, Baton Rouge
Abstract

During World War II, famous mathematician Abraham Wald invented an adaptive statistical inferential technique, Sequential Probability Ratio Tests (SPRTs), for testing the power of ammunition. In recent years of the information age, researchers of information technology have discovered that Wald's SPRTs can also be put into use for fighting against malicious intrusions of data networks. In this presentation, we attempt to illustrate, with minimum amount of mathematics, the working of SPRTs for the detection of network intrusions. Moreover, we will address the limitations of Wald's SPRTs and discuss possible new adaptive statistical inferential methods for improving the accuracy and speed of detection. Two common categories of network intrusions including, portscan attacks and Denial-of-Service attacks, are covered in our presentation.

When Tuesday, 17 April 2012, 14:00 - 15:00
Where Room 117 EE Building
More Hide Abstracts.   Announcement (PDF)
 
Title Introducing Control Flow into Vectorized Code
Speaker Jaewook Shin
Hewlett Packard
Abstract

Single instruction multiple data (SIMD) functional units are ubiquitous in modern microprocessors. Effective use of these SIMD functional units is essential in achieving the highest possible performance. Automatic generation of SIMD instructions in the presence of control flow is challenging, however, not only because SIMD code is hard to generate in the presence of arbitrarily complex control flow, but also because the SIMD code executing the instructions in all control paths may slow compared to the scalar original, which may bypass a large portion of the code. In this talk, I will present two techniques that improve on the previous approach. First, BOSCCs are generated in a nested fashion so that even BOSCCs themselves can be bypassed by other BOSCCs. Second, we generate all vec_any_* instructions to bypass even some predicate-defining instructions. On 14 kernels, the compiler achieves distinct speedups, including 1.99X over the previous technique that generates single-level BOSCCs and vec_any_ne only.

Bio

Dr. Jaewook Shin is a performance engineer at Hewlett Packard. He earned his Ph.D. degree from the University of Southern California in 2005 with his work on compiler optimizations for multimedia extension architectures. He had been a postdoctoral researcher and Enrico Fermi scholar in the Mathematics and Computer Science division at Argonne National Laboratory until he joined HP in 2010.

When Thursday, 19 April 2012, 15:30 - 16:30
Where Room 103 Design Building
More Hide Abstracts.   Announcement (PDF)
 
Title Speeding Up Nek5000 with Autotuning and Specialization
Speaker Jaewook Shin
Hewlett Packard
Abstract

Autotuning technology has emerged recently as a systematic process for evaluating alternative implementations of a computation, in order to select the best-performing solution for a particular architecture. Specialization optimizes code customized to a particular class of input data set. In this talk, I will talk about how compiler-based autotuning that incorporates specialization for expected data set sizes of key computations can be used to speed up Nek5000, a spectral-element code. Nek5000 makes heavy use of what are effectively Basic Linear Algebra Subroutine (BLAS) calls, but for very small matrices. Through autotuning and specialization, we can achieve significant performance gains over hand-tuned libraries (e.g., Goto, ATLAS, and ACML BLAS). We demonstrate more than 2.2X performance gains on an Opteron over the original manually tuned implementation, and speedups of up to 1.26X on the entire application running on 256 nodes of the Cray XT5 Jaguar system at Oak Ridge.

Bio

Dr. Jaewook Shin is a performance engineer at Hewlett Packard. He earned his Ph.D. degree from the University of Southern California in 2005 with his work on compiler optimizations for multimedia extension architectures. He had been a postdoctoral researcher and Enrico Fermi scholar in the Mathematics and Computer Science division at Argonne National Laboratory until he joined HP in 2010.

When Friday, 20 April 2012, 10:30 - 11:30
Where Room 338 Johnston Hall
More Hide Abstracts.   Announcement (PDF)