2015 IEEE High Performance
Extreme Computing Conference
(HPEC ‘15)
Nineteenth Annual HPEC Conference
15 - 17 September 2015
Westin Hotel, Waltham, MA USA
Bioinformatics & Big Data 3
3:00-4:40 in Eden Vale C1 - C2
Chair: Chansup Byun / MIT Lincoln Laboratory
Hierarchical Clustering and K-means Analysis of HPC Application Kernels Performance Characteristics
M.L. Grodowitz, Sarat Sreepathi, Oak Ridge National Lab
In this work, we present the characterization of a set of scientific kernels which are representative of the behavior of
fundamental and applied physics applications across a wide range of fields. We collect performance attributes in the form of
micro-operation mix and off-chip memory bandwidth measurements for these kernels. Using these measurements, we use two
clustering methodologies to show which applications behave similarly and to identify unexpected behaviors, without the need to
examine individual numeric results for all application runs. We define a methodology to combine metrics from various tools into a
single clustering visualization. We show that some kernels experience significant changes in behavior at varying thread counts
due to system features, and that these behavioral changes appear in the clustering analysis. We further show that application
phases can be analyzed using clustering to determine which section of an application is the largest contributor to behavioral
differences.
Multi-modal Sensor Registration for Vehicle Perception via Deep Neural Networks
Michael Giering, Vivek Venugopalan, Kishore Reddy, UTRC
The ability to simultaneously leverage multiple modes of sensor information is critical for perception of an automated vehicle's
physical surroundings. Spatio-temporal alignment of registration of the incoming information is often a prerequisite to analyzing
the fused data. The persistence and reliability of multi-modal registration is therefore the key to the stability of decision support
systems ingesting the fused information. LiDAR-video systems like on those many driverless cars are a common example of
where keeping the LiDAR and video channels registered to common physical features is important. We develop a deep learning
method that takes multiple channels of heterogeneous data, to detect the misalignment of the LiDAR-video inputs. A number of
variations were tested on the Ford LiDAR-video driving test data set and will be discussed. To the best of our knowledge the use
of multi-modal deep convolutional neural networks for dynamic real-time LiDAR-video registration has not been presented.
A Signals Processing and Big Data Framework for Monte Carlo Aircraft Encounters
Andrew Weinert, MIT Lincoln Laboratory
Developing a collision avoidance system that can meet safety standards required of commercial aviation is challenging. A
dynamic programming approach to collision avoidance has been developed to optimize and generate logics that are robust to
the complex dynamics of the national airspace. The current approach represents the aircraft avoidance problem as Markov
Decision Processes and independently optimizes a horizontal and vertical maneuver avoidance logics. This is a result of the
current memory requirements for each logic, simply combining the logics will result in a significantly larger representation. The
“curse of dimensionality” makes it computationally inefficient and infeasible to optimize this larger representation. However,
existing and future collision avoidance systems have mostly defined the decision process by hand. In response, a simulation-
based framework was built to better understand how each potential state quantifies the aircraft avoidance problem with regards
to safety and operational components. The framework leverages recent advances in signals processing and database, while
enabling the highest fidelity analysis of Monte Carlo aircraft encounter simulations to date. This framework enabled the
calculation of how well each state of the decision process quantifies the collision risk and the associated memory requirements.
Using this analysis, a collision avoidance logic that leverages both horizontal and vertical actions was built and optimized using
this simulation-based approach.
A Cloud-based approach to Big Graphs
Paul Burkhardt, Chris Waring, DoD
Thursday, September 17