Robust Parallel Adaptive Smoothing           Danny Gibbs II          Christopher K. Lee             CS205 Final Project   ...
Overview●   Adaptive Smoothing●   CPU and GPU/MPI Implementations●   Results●   Conclusions●   Summary
Adaptive Smoothing●   Independently smooths each pixel●   Preserves small scale features●   Smooths larger features over l...
Example of Adaptive Smoothing:    Threshold and MaxRad
Observation ID: 11759
CPU Implementation●   Traditional Chandra Observation    ●   dmimgadapt()    ●        Worst Case: O(n4)    ●   8192x8192 b...
GPU/MPI Implementation●   GPU    ●   Embarrassingly Parallel    ●   Three CUDA kernels    ●   Use full resolution image●  ...
Results
Conclusions●   CPU vs GPU    ●   Full Resolution Image        –  CPU: ~ 4.5 Days         – GPU: ~ 1.65 sec    ●        Spe...
Summary●   Adaptive Smoothing●   CPU and GPU/MPI Implementations●   Results●   Conclusions
AcknowledgmentsCIAO: Chandras data analysis systemFruscione et al. 2006, SPIE Proc. 6270, 62701V, D.R. Silvia & R.E. Doxse...
Upcoming SlideShare
Loading in …5
×

CS205 Final project

212 views

Published on

Published in: Technology, Business
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
212
On SlideShare
0
From Embeds
0
Number of Embeds
12
Actions
Shares
0
Downloads
1
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

CS205 Final project

  1. 1. Robust Parallel Adaptive Smoothing Danny Gibbs II Christopher K. Lee CS205 Final Project 4 December 2011
  2. 2. Overview● Adaptive Smoothing● CPU and GPU/MPI Implementations● Results● Conclusions● Summary
  3. 3. Adaptive Smoothing● Independently smooths each pixel● Preserves small scale features● Smooths larger features over larger area● Only scales to Threshold or MaxRad● Preserves energy (flux)
  4. 4. Example of Adaptive Smoothing: Threshold and MaxRad
  5. 5. Observation ID: 11759
  6. 6. CPU Implementation● Traditional Chandra Observation ● dmimgadapt() ● Worst Case: O(n4) ● 8192x8192 binned to 1024x1024 ● Run Time: ~ 30 minutes
  7. 7. GPU/MPI Implementation● GPU ● Embarrassingly Parallel ● Three CUDA kernels ● Use full resolution image● MPI ● Process Multiple Observations
  8. 8. Results
  9. 9. Conclusions● CPU vs GPU ● Full Resolution Image – CPU: ~ 4.5 Days – GPU: ~ 1.65 sec ● Speed Up of ~105
  10. 10. Summary● Adaptive Smoothing● CPU and GPU/MPI Implementations● Results● Conclusions
  11. 11. AcknowledgmentsCIAO: Chandras data analysis systemFruscione et al. 2006, SPIE Proc. 6270, 62701V, D.R. Silvia & R.E. Doxsey,eds.Special thanks to Kenny Glotfelty of the Harvard-Smithsonian Center for Astrophysics

×