Poster Gallery

Explore A Broad Range Of GPU Posters

  Explore over 140 research posters detailing how the brightest minds from a range of topics and industries are accelerating applications with the power of GPUs. Please sign up to view high resolution PDF.

  •

    Accelerated Data Science
    GPUDirect for Billion Rows per Second Data Processing on Single-Node PostgreSQL
    Manifold Regularization For Time Series Embedding
    CheetahDB: A System for High-Throughput Database Processing on GPUs

    CheetahDB: A System for High-Throughput Database Processing on GPUs

    Chengcheng Mou, University of South Florida

    OpenData Anonymization with Generative Adversarial Networks and Differential Privacy
    Optimizing Work Scheduling and Memory Usage for Complex Database Queries on GPUs

  • Algorithms / Numerical Techniques
    A New Convolution Algorithm to Leverage Tensor Cores

    A New Convolution Algorithm to Leverage Tensor Cores

    Mickaël Seznec, Thales / Centralesupelec

    GPU Parallel Computing Implementation of the Height Digital Image Correlation (hDIC) Technique
    Dramatic Acceleration of Quantum Transport Simulations: Solving Non-Equilibrium Green’s Function with GPU Devices
    Tensor Core Accelerated Sparse GEMM

    Tensor Core Accelerated Sparse GEMM

    Orestis Zachariadis, University of Cordoba

     

    Self-Supervised Generative Adversarial Networks Compression
    OmniTrack: Real-Time Detection and Tracking of Objects, Text, and Logos in Video
    Efficient 3D Convolutional Network Design for Human Instance-Level Video Action Recognition

    Efficient 3D Convolutional Network Design for Human Instance-Level Video Action Recognition

    Inwoong Lee, Artificial Intelligence Research Institute (AIRI)

    A Software Systolic Array on GPUs

    A Software Systolic Array on GPUs

    Mohamed Wahib, AIST-Tokyo Tech Real World Big-Data Computation Open Innovation Laboratory, National Institute of Advanced Industrial Science and Technology  

    GraphDefense: Toward Robust Large-Scale Graph Convolutional Network

    GraphDefense: Toward Robust Large-Scale Graph Convolutional Network

    Xiaoyun Wang, University of California, Davis

    Fromage optimiser for deep neural networks

    Fromage optimiser for deep neural networks

    Jeremy Bernstein, California Institute of Technology

    Hybrid Option Pricing through AI and GPU-Powered SDEs Solvers

    Hybrid Option Pricing through AI and GPU-Powered SDEs Solvers

    Pawel Przybyłowicz, AGH University of Science and Technology

    GPU-Accelerated SPICE for Deep Submicron Analog Simulation

  • Autonomous Machines
    Deep Learning-Based Anti-Drone System
    MemEAPF for Mobile Robot Path Planning on Jetson Platform

    MemEAPF for Mobile Robot Path Planning on Jetson Platform

    Ulises Orozco-Rosas, CETYS Universidad  

    Sim-to-Real: Virtual Guidance for Robot Navigation

    Sim-to-Real: Virtual Guidance for Robot Navigation

    Chun-Yi Lee, National Tsing Hua University

    Mobile Manipulation Toward Cleaning Dining Tables

    Mobile Manipulation Toward Cleaning Dining Tables

    Ka-Shing Chung, National University of Singapore

    Beyond-Line-of-Sight (BLOS) Perception System for Autonomous Vehicles

    Beyond-Line-of-Sight (BLOS) Perception System for Autonomous Vehicles

    Chanyoung Chung, KAIST (Korea Advanced Institute of Science and Technology)

    Geolocation of Traffic Lights and Signs Using Dashcam: Toward Low-Cost Map Maintenance
    Drones as a Tool for Development and Safety Validation of Automated Driving Functions
    End-to-End Learning with Combined
    Improving Performance Limits of Obstacle-Avoidance Driving with Randomized Model Predictive Control Using GPU
    Temporal Information Prediction for Perception in Autonomous Vehicles

  • Climate / Weather / Ocean Modeling
    Improving Ocean Model Framework NEMO with GPU Port and Asynchronous Execution
    Evaluation of a Multi-GPU Optimized Non-Hydrostatic Ocean Model with Multigrid Preconditioned Conjugate Gradient Method
    AceCAST GPU-Enabled Weather Research and Forecasting Model Development and Applications

  • Computational Fluid Dynamics
    Toward Optimal Implementation of Lattice Boltzmann CFD Simulator for Multi-GPU Clusters

    Toward Optimal Implementation of Lattice Boltzmann CFD Simulator for Multi-GPU Clusters

    Jakub Klinkovský, Department of Mathematics, Faculty of Nuclear Sciences and Physical Engineering, Czech Technical University, Prague

    An Improved Immersed Boundary-Lattice Boltzmann Method for Incompressible Fluid-Flow Simulations on GPU

    An Improved Immersed Boundary-Lattice Boltzmann Method for Incompressible Fluid-Flow Simulations on GPU

    Pavel Eichler, Faculty of Nuclear Sciences and Physical Engineering, Czech Technical University, Prague

    Performance Comparison of Search for Neighbor-Particle in MPS on Xeon, Xeon Phi and GPU

    Performance Comparison of Search for Neighbor-Particle in MPS on Xeon, Xeon Phi and GPU

    Takaaki Miyajima, RIKEN Center for Computational Science

    Toward Large-Scale Steady-State Computational Fluid Dynamics Acceleration with Point Cloud Networks

  • Computational Physics
    N-Body Simulation of Binary Star Mass Transfer

    N-Body Simulation of Binary Star Mass Transfer

    Mason McCallum, Tarleton State University

    Explosive DGX Performance for Weapon Component Modeling

    Explosive DGX Performance for Weapon Component Modeling

    Dan Ibanez, Sandia National Laboratories

    Deep-Learning Model for Finding New Superconductors

    Deep-Learning Model for Finding New Superconductors

    Tomohiko Konno, National Institute of Information and Communications Technology

    EXtended Particle System (XPS): High-Performance Particle Simulation

  • Deep Learning & Data Science
    AI-Assisted Cell Tower Inspection with Drones and GPUs
    Is the Label Trustworthy? Training Better Deep-Learning Models via Uncertainty Mining Net
    CapsNet-Lite: A Lightweight and High Performance CapsNet Architecture

    CapsNet-Lite: A Lightweight and High Performance CapsNet Architecture

    Francisco José García, Universidad Rey Juan Carlos

    Bike-Share Demand Prediction from Partial Future data Using Conditional Variational AutoEncoders
    BasketballGAN : Generating Basketball Play Simulation Through Sketching
    Multi-Task Learning for Sparse Sensor Body Tracking

    Multi-Task Learning for Sparse Sensor Body Tracking

    Aditya Tewari, Xsens Technologies

    Adversarial Learning of Deepfakes in Financial Accounting

    Adversarial Learning of Deepfakes in Financial Accounting

    Marco Schreyer, University of St. Gallen

    Representing Documents Using Contextualized Embedding from Transformers
    Recipes for Explaining Text Classification Models
    Intelligent Risk Scenario Creation

    Intelligent Risk Scenario Creation

    Amit Kalele, Tata Consultancy Services Limited

    Learning Human Objectives by Evaluating Hypothetical Behavior

    Learning Human Objectives by Evaluating Hypothetical Behavior

    Siddharth Reddy, University of California at Berkeley

    Trajectory-Wise Control Variates for Variance Reduction in Policy Gradient Methods
    How Humans Leverage Machines for Quick Response in High-Demand Customer Service Environment

    How Humans Leverage Machines for Quick Response in High-Demand Customer Service Environment

    Weicheng Liu, Wells Fargo

    Deep Autoencoder-Based Statistical Arbitrage During a Bearish Market

  • Deep Learning Training at Scale
    Fast Data Pre-Processing with NVIDIA Data Loading Library (DALI)
    Using Robust Networks to Inform Lightweight Models in Semi-Supervised Learning for Object Detection
    Extremely Accelerated Deep Learning: ResNet-50 Training in 70.4 Seconds
    GradZip: Gradient Compression Using Alternating Matrix Factorization for Large-scale Deep Learning
    Knowledge Transfer Graph for Deep Collaborative Learning

    Knowledge Transfer Graph for Deep Collaborative Learning

    Hironobu Fujiyoshi, Chubu University

  •

    Design & Engineering
    N-body Adaptive Optimization of Lattice Towers

    N-body Adaptive Optimization of Lattice Towers

    Jaryd Domine, Tarleton State University

    Semi-Supervised Three-Dimensional Reconstruction with Generative Adversarial Networks
    ESD Risk Mitigation Using CUDA-Accelerated Algorithm for MOSFET Connectivity Checking
    cGAN for Automatic Site Selection of Cultural Venues

  • Frameworks / Libraries / Runtimes
    Acceleration of Test Data Quality Assurance Technology Using Neuron Coverage
    Accelerating Large Seismic Simulation Code with PACC Framework
    D-GEN: A Deep Learning Data Generation Framework for Artificial Intelligence

  • Genomics
    Rapid Pathogen Genomics using Nanopore Sequencing and GPUs

    Rapid Pathogen Genomics using Nanopore Sequencing and GPUs

    Devin Drown, University of Alaska Fairbanks

  •

    HPC & AI
    MolecularRNN: Generating Realistic Molecular Graphs with Optimized Properties
    AI-Solver: Deep Learning-Based Platform for Real-time Simulations on Structured (Grid) and Unstructured (Mesh) Data Representations
    Turbulence Forecasting via Neural ODEs

    Turbulence Forecasting via Neural ODEs

    Gavin Portwood, Los Alamos National Laboratory

    Reynolds-Averaged Turbulence Modeling Using Neural Networks with Grid Information

    Reynolds-Averaged Turbulence Modeling Using Neural Networks with Grid Information

    Huiying Ren, Computer Network Information Center, Chinese Academy of Sciences

    Container-Based Artificial Intelligence Applications Deployment Platform for GPU High Performance Computing Clusters

    Container-Based Artificial Intelligence Applications Deployment Platform for GPU High Performance Computing Clusters

    Rongqiang Cao, Computer Network Information Center, Chinese Academy of Sciences

    Enhancing Intra-Node Multi-GPU Stencil Calculations on DGX-2 Using Concurrent-Addressing with Unified Memory
    TNL: Template Numerical Library for Modern Parallel Architectures

    TNL: Template Numerical Library for Modern Parallel Architectures

    Tomáš Oberhuber, Czech Technical University in Prague

    Runtime Analysis of Spatial Structure: A CUDA Implementation of Minkowski Functionals
    Heterocomputing and Transprecision Computing in Unstructured Low-Order Finite-Element Analyses on Volta GPUs
    Low-Cost OpenACC Porting of Matrix Solver with FP21-FP32-FP64 Computing: an Earthquake Application
    Data Mining Pipeline for Predictive Synthesis of Advanced Materials

    Data Mining Pipeline for Predictive Synthesis of Advanced Materials

    Olga Kononova, University of California, Berkeley

    Applications of Convolutional Neural Network to the Important Earth Science Problems

    Applications of Convolutional Neural Network to the Important Earth Science Problems

    Daisuke Sugiyama, Japan Agency for Marine-Earth Science and Technology

    Multiphysics Software Development in the Age of AI
    Implementation of Artificial Intelligence (AI)/Deep Learning Disruption Predictor into a Plasma Control System

    Implementation of Artificial Intelligence (AI)/Deep Learning Disruption Predictor into a Plasma Control System

    Ge Dong, Princeton Plasma Physics Laboratory

  •

    Inference, Optimization, & Deployment
    Tiki: A Prototype Virtual Assistant for VQA

    Tiki: A Prototype Virtual Assistant for VQA

    Brent Biseda, Seneca Resources

    Benanza: Automatic μBenchmark Generation to Compute ”Lower-bound” Latency and Inform Optimizations of Deep Learning Models on GPUs
    AWDF: An Adaptive Weighted Deep Fusion Architecture for Multi-Modality Learning
    Automated and Intelligent Cloud Bursting Tool using Deep Neural Networks