Suren Chilingaryan, Andreas Kopmann

Slides:



Advertisements
Similar presentations
DEVELOPMENT OF ONLINE EVENT SELECTION IN CBM DEVELOPMENT OF ONLINE EVENT SELECTION IN CBM I. Kisel (for CBM Collaboration) I. Kisel (for CBM Collaboration)
Advertisements

Tango Meeting DESY Status Report Thorsten Kracht Grenoble, 13. May 2009.
Multi-core and tera- scale computing A short overview of benefits and challenges CSC 2007 Andrzej Nowak, CERN
HPCC Mid-Morning Break High Performance Computing on a GPU cluster Dirk Colbry, Ph.D. Research Specialist Institute for Cyber Enabled Discovery.
A many-core GPU architecture.. Price, performance, and evolution.
GPU Computing with CUDA as a focus Christie Donovan.
Multi Agent Simulation and its optimization over parallel architecture using CUDA™ Abdur Rahman and Bilal Khan NEDUET(Department Of Computer and Information.
DCABES 2009 China University Of Geosciences 1 The Parallel Models of Coronal Polarization Brightness Calculation Jiang Wenqian.
Introduction What is GPU? It is a processor optimized for 2D/3D graphics, video, visual computing, and display. It is highly parallel, highly multithreaded.
GPGPU overview. Graphics Processing Unit (GPU) GPU is the chip in computer video cards, PS3, Xbox, etc – Designed to realize the 3D graphics pipeline.
GPGPU platforms GP - General Purpose computation using GPU
HPCC Mid-Morning Break Dirk Colbry, Ph.D. Research Specialist Institute for Cyber Enabled Discovery Introduction to the new GPU (GFX) cluster.
Digital Graphics and Computers. Hardware and Software Working with graphic images requires suitable hardware and software to produce the best results.
High Performance Computing G Burton – ICG – Oct12 – v1.1 1.
MACHINE VISION GROUP Graphics hardware accelerated panorama builder for mobile phones Miguel Bordallo López*, Jari Hannuksela*, Olli Silvén* and Markku.
CuMAPz: A Tool to Analyze Memory Access Patterns in CUDA
Computer Graphics Graphics Hardware
BY: ALI AJORIAN ISFAHAN UNIVERSITY OF TECHNOLOGY 2012 GPU Architecture 1.
By Arun Bhandari Course: HPC Date: 01/28/12. GPU (Graphics Processing Unit) High performance many core processors Only used to accelerate certain parts.
Accelerating image recognition on mobile devices using GPGPU
GPU Architecture and Programming
Graphics Card Andrew Kasper MYP 5.
GPUs – Graphics Processing Units Applications in Graphics Processing and Beyond COSC 3P93 – Parallel ComputingMatt Peskett.
Contemporary Languages in Parallel Computing Raymond Hummel.
From Turing Machine to Global Illumination Chun-Fa Chang National Taiwan Normal University.
GPGPU introduction. Why is GPU in the picture Seeking exa-scale computing platform Minimize power per operation. – Power is directly correlated to the.
Calculator in a Box Co. Lead designer: Bradley Phelps.
Large-scale geophysical electromagnetic imaging and modeling on graphical processing units Michael Commer (LBNL) Filipe R. N. C. Maia (LBNL-NERSC) Gregory.
S. Pardi Frascati, 2012 March GPGPU Evaluation – First experiences in Napoli Silvio Pardi.
KIT – University of the State of Baden-Wuerttemberg and National Research Center of the Helmholtz Association Institute for Data Processing and Electronics.
NVIDIA® TESLA™ GPU Based Super Computer By : Adam Powell Student # For COSC 3P93.
Heterogeneous Processing KYLE ADAMSKI. Overview What is heterogeneous processing? Why it is necessary Issues with heterogeneity CPU’s vs. GPU’s Heterogeneous.
Sobolev(+Node 6, 7) Showcase +K20m GPU Accelerator.
The Limits of Volunteer Computing Dr. David P. Anderson University of California, Berkeley March 20, 2011.
“SMT Capable CPU-GPU Systems for Big Data”
Computer Graphics Graphics Hardware
Transformer for your computer
GPU Architecture and Its Application
Kai Li, Allen D. Malony, Sameer Shende, Robert Bell
Generalized and Hybrid Fast-ICA Implementation using GPU
Graphics Processor Graphics Processing Unit
Stencil-based Discrete Gradient Transform Using
Parallel Plasma Equilibrium Reconstruction Using GPU
Our Graphics Environment
CS427 Multicore Architecture and Parallel Computing
GPU Computing Jan Just Keijser Nikhef Jamboree, Utrecht
Get more done with Windows 10 Pro for Workstations
Introduction to Computer Graphics
What is GPU? how does it work?
Enabling machine learning in embedded systems
tomography: from medicine to accelerator physics, and back again
Constructing a system with multiple computers or processors
Texas Instruments TDA2x and Vision SDK
Introduction to Parallelism.
Multi-Layer Perceptron On A GPU
From Turing Machine to Global Illumination
Scientific computing in x-ray microscopy
Unit 2 Computer Systems HND in Computing and Systems Development
The Yin and Yang of Processing Data Warehousing Queries on GPUs
MASS CUDA Performance Analysis and Improvement
NVIDIA Fermi Architecture
Constructing a system with multiple computers or processors
Constructing a system with multiple computers or processors
About Hardware Optimization in Midas SW
The Free Lunch Ended 7 Years Ago
Constructing a system with multiple computers or processors
1.1 The Characteristics of Contemporary Processors, Input, Output and Storage Devices Types of Processors.
Computer Graphics Graphics Hardware
BWLOCK++: Protecting GPU Kernels on Integrated CPU-GPU Platforms
Graphics Processing Unit
Presentation transcript:

A Novel Approach for Online-Monitoring for High Data-Rate Image-Based Instrumentation Suren Chilingaryan, Andreas Kopmann Forschungszentrum Karlsruhe, Germany Goal: Processing of large data sets Data set 10-50GB Speed up hours → < 1 minute Solution: Standard graphic adapter Universal programming (CUDA, OpenCL) Teraflop performance High bandwidth Benefit: Improve experiment quality Experiment control Tomographic reconstruction 3D Imaging applications produce large data sets. Several pictures from different angles or during manipulation are taken and the 3D object needs to be reconstructed from the image sequence. The size of the data sets is typically in the order of 10-50GB. Regular PCs need for this task several hours up to one day of computation time – this is too long for direct feed back (quality of the probe, control of the experiment). Even the transport of the data set takes too long too come to close real time results. Utilizing standard graphics adapters can provide a solution. Actual graphic adapters provide a multiple of the computation power of actual desktop CPUs. The adapter is connect with high bandwidth (PCI Express). And all manufactures offer general purpose libraries that allow to use the graphics adapters also for non graphics purposes. In order to have a general programming interface with OpenCL a new standard for GPU, DSP and CPU systems is developed. The first results for image processing applications show that quite large performance increase can be reached. It seems to be possible to come in the desired regions of realtime or near-realtime monitoring. Of course depending on the application and the resolution of the images. Further planned applications are online-simulation for partical physics. Here simulations are more and more needed in parallel with the experiments. Examples: Fieldline simulations for KATRIN neutrino experiment or cosmic ray simulations. + Graphic co-processors offer multi teraflops in a single PC KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) 1 1

Application: Strain Measurement Data Set: Number of Images: 11 (521) Resolution: 2208x3000 Total size: 200 MB (10GB) Results: CPU reconstruction: ~10 hours (16 cores) GPU reconstruction: 30 min Image corr / All Unoptimized 173 212 Software Optimization 97 140 CUDA/FFT plugin 60 104 FULL CUDA support 9 45 Parallel Memory Transfers 6 42 Parallel Computations 6 25 Nvidia Tesla S1070 2 12

Application: X-ray Tomography PyHST - High Speed Tomography Reconstruction ESRF (European Synchrotron Radiation Facility) Polygone Scientifique Louis Néel, 6 rue Jules Horowitz, 38000 GRENOBLE Data Set: Number of Images: 2000 Resolution: 1776x1707 Total size: 24 GB Results: CPU reconstruction: ~10 hours (16 cores) GPU reconstruction: 30 min Porose polyethylene grains in a conical plastic holder 3 | Dr. Andreas Kopmann | PNI Bonus Programm 27.3.2009