GPU activities at FI MUNI and their results Ji Matela, Ji Filipovi - - PowerPoint PPT Presentation

▶
gpu activities at fi muni and their results
SMART_READER_LITE
LIVE PREVIEW

GPU activities at FI MUNI and their results Ji Matela, Ji Filipovi - - PowerPoint PPT Presentation

Dec 09, 2022 460 likes •580 views

GPU activities at FI MUNI and their results Ji Matela, Ji Filipovi <matela@ics.muni.cz> , <fila@ics.muni.cz> Laboratory of Advanced Network Technologies MetaCentrum CESNET Grid Computing Seminar 2010 Praha,

slide-1
SLIDE 1

1/11

GPU activities at FI MUNI and their results

Jiří Matela, Jiří Filipovič

<matela@ics.muni.cz>, <fila@ics.muni.cz> Laboratory of Advanced Network Technologies MetaCentrum CESNET

Grid Computing Seminar 2010 Praha, 2010–10–15

slide-2
SLIDE 2

2/11

AutoGrid

  • Potential maps generation for molecular docking
  • The most computationally expensive parts accelerated on

GPU

  • CPU part analyzed and modified
slide-3
SLIDE 3

3/11

AutoGrid – Speedup

Accelerated design shows speedup of up to 400×

50 100 150 200 250 300 350 400 450 50 100 150 200 250 300 350 speedup grid size disntance-dependent constant

slide-4
SLIDE 4

4/11

Discrete Wavelet Transform (DWT)

  • Digital signal processing technique
  • Application in diverse areas
  • digital speech recognition
  • multi-resolution video processing
  • data compression
slide-5
SLIDE 5

5/11

DWT – Speedup

Our GPU implementation shows 68× speedup

slide-6
SLIDE 6

6/11

Real-Time Video and Fast Large-Scale Image Compression

  • Ongoing project
  • Real-time compression and transmission of video in HD

post-HD resolutions

  • Fast compression of pathological images of resolutions in
  • rder of gigapixels
  • GPU acceleration of JPEG2000
slide-7
SLIDE 7

7/11

JPEG2000 compression process

DWT Context Modeling Arithmetic Encoding Data compression

slide-8
SLIDE 8

8/11

Context-Modeling in JPEG2000

  • Serial algorithm
  • Redesign to fit specifics of GPUs
  • 12× faster compared to JasPer CPU implementation
slide-9
SLIDE 9

9/11

Soft tissues simulations

  • Haptic surgical simulators
  • Simulations modelled using Finite Element Method

(FEM)

  • FEM discretizes the modeled object as a mesh of elements
  • Per element computation and system of equations solving
  • Per element computation is complex problem so that it

needs to be decomposed into several GPU functions

  • Not easy to choose decomposition granularity
  • Manual development of as small functions as possible
  • Automatic fusion into lager functions
slide-10
SLIDE 10

10/11

Preliminary performance gain of fusion

80% gain compared to non fused approach

2000 4000 6000 8000 10000 10000 20000 30000 40000 50000 60000 thousands elements/s # elements GPU SMEM/GMEM GPU GMEM CPU 1 core

slide-11
SLIDE 11

11/11

Thank you for you attention!

Q?/A! <matela@ics.muni.cz>, <fila@ics.muni.cz> http://www.sitola.cz/