BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260522T150153Z
LOCATION:C2/3/4 Ballroom
DTSTART;TZID=America/Chicago:20181115T083000
DTEND;TZID=America/Chicago:20181115T170000
UID:submissions.supercomputing.org_SC18_sess327@linklings.com
SUMMARY:ACM Student Research Competition Posters
DESCRIPTION:SC18 ACM Student Research Competition Posters will be on displ
 ay on Tuesday, Wednesday, Thursday from 8:30am to 5pm in the C2/3/4 Ballro
 om.\n\nAccelerating Big Data Processing in the Cloud with Scalable Communi
 cation and I/O Schemes\n\nWith the advent of cloud computing, the field of
  Big Data has seen rapid growth. Most cloud providers provide hardware res
 ources such as NVMe SSDs, large memory nodes, and SR-IOV. This opens up th
 e possibility of large-scale high-performance data analytics and provides 
 opportunities to use these res...\n\n\nShashank Gugnani (Ohio State Univer
 sity)\n---------------------\nModeling Single-Source Shortest Path Algorit
 hm Dynamics to Control Performance and Power Tradeoffs\n\nThis work presen
 ts a new methodology to improve the performance of parallel algorithms by 
 tuning the amount of available parallelism for execution throughout the ru
 ntime. As such, we expose key parameters controlling the performance and p
 arallelism of the algorithm and build a software-based control...\n\n\nSar
 a Karamati, Jeffrey Young, and Rich Vuduc (Georgia Institute of Technology
 )\n---------------------\nMeasuring Swampiness: Quantifying Chaos in Large
  Heterogeneous Data Repositories\n\nAs scientific data repositories and fi
 lesystems grow in size and complexity, they become increasingly disorganiz
 ed. The coupling of massive quantities of data with poor organization make
 s it challenging for scientists to locate and utilize relevant data, thus 
 slowing the process of analyzing data of...\n\n\nLuann C. Jung (Massachuse
 tts Institute of Technology, University of Chicago) and Brendan T. Whitake
 r (Ohio State University, University of Chicago)\n---------------------\nH
 ardware Transactional Persistent Memory\n\nThis research solves the proble
 m of creating durable transactions in byte-addressable Non-Volatile Memory
  or Persistent Memory (PM) when using Hardware Transactional Memory (HTM)-
 based concurrency control.  It shows how HTM transactions can be ordered c
 orrectly and atomically into PM by the use of a...\n\n\nEllis Giles (Rice 
 University)\n---------------------\nAccelerating DNA Long Read Mapping wit
 h Emerging Technologies\n\nDNA sequencing technologies output only short f
 ragments of a genome, called reads. New single-molecule real-time sequenci
 ng technologies can produce long reads, up to tens of thousands base pairs
 , within minutes. However, these long reads may contain up to 15% errors.\
 n\nTo construct a genome from DNA...\n\n\nRoman Kaplan (Israel Institute o
 f Technology)\n---------------------\nOoO Instruction Benchmarking Framewo
 rk on the Back of Dragons\n\nIn order to construct an accurate instruction
  execution model for modern out-of-order micro architectures, an accurate 
 description of instruction latency, throughput and concurrency is indispen
 sable. Already existing resources and vendor provided information is neith
 er complete nor detailed enough a...\n\n\nJulian Hammer (University of Erl
 angen-Nuremberg, RRZE)\n---------------------\nPotC: Many-Body Potential I
 mplementations à La Carte\n\nMolecular dynamics is a valuable investigatio
 n tool for simulations in computational chemistry and materials science. I
 n these simulations, atoms move according to so-called potentials, functio
 ns that typically describe the distance-dependent interactions between pai
 rs of atoms. For some application...\n\n\nMarkus Höhnerbach (RWTH Aachen U
 niversity)\n---------------------\nNumerical Simulation of a Flue Instrume
 nt with Finite-Difference Lattice Boltzmann Method using GPGPU\n\nIn this 
 work, we discuss the possibility of using GPGPU techniques for Aeroacousti
 c Simulation (especially for flue instruments) with the finite-difference 
 lattice Boltzmann method (FDLBM).   Compressible flow simulation has been 
 used in direct aeroacoustic simulation; however, the computational cos...\
 n\n\nRyoya Tabata (Kyushu Institute of Technology)\n---------------------\
 nDesigning Shared Address Space MPI Libraries in Many-Core Era\n\nThe emer
 gence of modern multi-/many-cores has put more emphasis on optimizing intr
 a-node communication. Existing designs in MPI libraries that work on the c
 oncept of distributed address spaces incur the overhead of intermediate me
 mory copies to stage the data between processes. This can lead to seve...\
 n\n\nJahanzeb Maqbool Hashmi (Ohio State University)\n--------------------
 -\nIdentifying Network Data Transfer Bottlenecks in HPC Systems\n\nImprovi
 ng network data transfer performance is a major factor for improving high 
 performance computing systems. Most studies analyze data transfer and file
  system IO performance separately, but understanding the relationship betw
 een the two is essential for optimizing scheduling and resource managem...
 \n\n\nKaren Tu (Lawrence Berkeley National Laboratory; University of Calif
 ornia, Berkeley)\n---------------------\nStudying the Impact of Power Capp
 ing on MapReduce-Based, Data-Intensive Mini-Applications on Intel KNL and 
 KNM Architectures\n\nIn this poster, we quantitatively measure the impacts
  of data movement on performance in MapReduce-based applications when exec
 uted on HPC systems. We leverage the PAPI ‘powercap’ component to identify
  ideal conditions for execution of our applications in terms of (1) datase
 t characteristics (i.e., ...\n\n\nJoshua H. Davis (University of Delaware)
 \n---------------------\nUsing Integrated Processor Graphics to Accelerate
  Concurrent Data and Index Structures\n\nWith the advent of computing syst
 ems with on-die integrated processor graphics (iGPU), new programming chal
 lenges have emerged from these heterogeneous systems. We proposed differen
 t data and index structure algorithms that can benefit from the Intel's iG
 PU architecture and the C for Media (CM) prog...\n\n\nJoel Fuentes (Univer
 sity of California, Irvine)\n---------------------\nSupercomputing for the
  Multi-Driver Routing\n\nSupercomputing is essential for routing traffic b
 y providing drivers the optimal routes with minimal traveling distances or
  time. The unique challenges that require supercomputers to overcome are o
 f multiple folds: numerous drivers, massive simultaneous requests, multipl
 e locations, and needs of ins...\n\n\nZeyang Ye (Stony Brook University)\n
 ---------------------\nHolistic Root Cause Analysis of Node Failures in Pr
 oduction HPC\n\nProduction HPC clusters endure failures incurring computat
 ion and resource wastage. Despite the presence of various failure detectio
 n and prediction schemes, a comprehensive understanding of how nodes fail 
 considering various components and layers of the system is required for su
 stained resilience. ...\n\n\nAnwesha Das (North Carolina State University)
 \n---------------------\nRecursive Algebraic Coloring Engine\n\nMany itera
 tive numerical methods for sparse systems and building blocks of sparse li
 near algebra are difficult to parallelize due to data dependencies. These 
 may be loop-carried dependencies as they occur in solvers like Gauss-Seide
 l or write conflicts as in symmetric sparse matrix vector.  Most of ...\n\
 n\nChristie Louis Alappat (University of Erlangen-Nuremberg)\n------------
 ---------\nMitigating Performance and Progress Variability in Iterative As
 ynchronous Algorithms\n\nLarge HPC machines are susceptible to irregular p
 erformance. Factors like chip manufacturing differences, heat management, 
 and network congestion combine to result in varying execution time for the
  same code and input sets. Asynchronous algorithms offer a partial solutio
 n. In these algorithms, fast ...\n\n\nJusts Zarins (University of Edinburg
 h)\n---------------------\nNautDB: Toward a Hybrid Runtime for Processing 
 Compiled Queries\n\nGeneral purpose operating and database system suffer u
 nder the load of their generality which makes achieving optimal performanc
 e extremely hard, especially on modern hardware. The goal of this research
  is to integrate, for the first time, specialization techniques from the O
 S community (hybrid runt...\n\n\nSamuel Grayson (University of Texas, Dall
 as)\n---------------------\nAccelerating 2D FFT: Exploit GPU Tensor Cores 
 through Mixed-Precision\n\nThe two-dimensional Fourier Transform is a wide
 ly-used computational kernel in many HPC applications. The popular NVIDIA 
 cuFFT library provides a simple interface to compute 2D FFT on GPUs, but i
 t's yet to utilize the recent hardware advancement in half-precision float
 ing-point arithmetic. In this p...\n\n\nXiaohe Cheng (Hong Kong University
  of Science and Technology) and Anumeena Sorna (National Institute of Tech
 nology, Tiruchirappalli)\n---------------------\nSimFS: A Simulation Data 
 Virtualizing File System Interface\n\nIn the big (simulation) data era, si
 mulations often produce petabytes of data to be stored in parallel filesys
 tems or large-scale databases. This data is accessed, often by thousands o
 f analysts and scientists, over the course of decades. However, storing th
 ese volumes of data for long time periods ...\n\n\nSalvatore Di Girolamo (
 ETH Zurich)\n---------------------\nDendro-GR: Massively Parallel Simulati
 ons of Binary Black Hole Intermediate-Mass-Ratio Inspirals\n\nWe present a
  portable and highly-scalable algorithm and framework that targets problem
 s in the astrophysics and numerical relativity communities. This framework
  combines together a parallel octree-refined adaptive mesh with wavelet ad
 aptive multiresolution and a physics module to solve the Einstein ...\n\n\
 nMilinda Fernando (University of Utah)\n---------------------\nGeomancy: A
 utomated Data Placement Optimization\n\nExascale cloud storage and High-Pe
 rformance Computing Systems (HPC) deliver unprecedented storage capacity a
 nd levels of computing power, though the full potential of these systems r
 emain untapped because of inefficient data placement. Changes in data acce
 ss patterns can cause a system's performance...\n\n\nOceane Bel (Universit
 y of California, Santa Cruz)\n---------------------\nPrecomputing Outputs 
 of Hidden Layers to Speed Up Deep Neural Network Training\n\nDeep learning
  has recently emerged as a powerful technique for many tasks including ima
 ge classification. A key bottleneck of deep learning is that the training 
 phase takes a lot of time, since state-of-the-art deep neural networks hav
 e millions of parameters and hundreds of hidden layers. The early...\n\n\n
 Sohil Lal Shrestha (University of Texas, Arlington)\n---------------------
 \nEulerian Algorithms for the Discretization of Plasma Kinetic Equations\n
 \nWhile fluid models are common tools in the study of plasmas, many of the
 se systems, whether in astrophysics or the lab, are only weakly collisiona
 l and far from equilibrium, making them more accurately described by kinet
 ic equations. Kinetic equations can be computationally demanding due to th
 e need...\n\n\nJames L. Juno (University of Maryland)\n-------------------
 --\nAccelerating Microscope Data Analysis Using Parallel Computing\n\nSing
 le-Molecule Localization Microscopy (SMLM) techniques deal with the diffra
 ction limit of fluorescent microscopy by localizing single molecules with 
 high precision by stochastically switching molecules on and off. Thousands
  of camera frames containing subsets of blinking molecules are recorded to
 ...\n\n\nJohn Ravi (North Carolina State University)\n--------------------
 -\nMonitoring Parsl Workflows\n\nAs a Python library that enables workflow
 s, Parsl gives users the ability to define complex workflows in Python and
  run them in parallel on any computer system. This poster describe the pro
 cess of adding monitoring to Parsl. Simple and comprehensive monitoring of
  a workflow’s state and resource usag...\n\n\nConnor Pigg (University of I
 llinois)\n\nRegistration Category: Tech Program Reg Pass, Exhibits Reg Pas
 s
END:VEVENT
END:VCALENDAR