Presentation is loading. Please wait.

Presentation is loading. Please wait.

Computation Time Analysis - Climate Reanalysis Data Dipanwita Dasgupta University of Notre Dame Graduate Operating Systems.

Similar presentations


Presentation on theme: "Computation Time Analysis - Climate Reanalysis Data Dipanwita Dasgupta University of Notre Dame Graduate Operating Systems."— Presentation transcript:

1 Computation Time Analysis - Climate Reanalysis Data Dipanwita Dasgupta University of Notre Dame Graduate Operating Systems

2 Motivation Climate Analysis : Why it is important?  Increase in occurrence of climate hazards Climate Reanalysis Data  Data Centric Approach  Climate Network Slide 2

3 Dataset National Centre for Environemental Prediction / National Centre for Atmospheric Research (NCEP/NCAR) Reanalysis Dataset  Composed of data at 17 pressure levels  Total of approximately 10000 grid points  Factors affecting climate Slide 3

4 Background Climate Network Model  Limited to use 7 factors affecting climate  Affects the predictive modeling  Computation Time Slide 4 out of x

5 Problem Computation has 3 steps 1. Reading the data from file 2. Calculation at each level 3. Combining the results Step 2 – highly computation intensive The present code can only handle 20 units of data at a time Slide 5

6 Slide 6

7 Actual Work Analyzed time taken to run on a single machine Distributed Framework  Steps 1 and 2 mentioned in previous slide for each level are independent of each other  Ran in a distributed fashion  Used the CRC SGE Machine Slide 7

8 Assumptions Used only one parameter – Geopotential Height Only one measure of dispersion – Euclidean Distance Processing is similar for other parameters as well as for measures of dispersion Slide 8

9 Experimental Set-up NCEP Reanalysis Dataset 20 units of longitude Sequential Execution  Used the school workstation desktop Distributed Framework  Used opteron.crc.nd.edu Slide 9

10 Distributed Framework: Setup opteron.crc.nd.edu Submitted Bash script Ran 10 simulations per level Took the average Slide 10

11 Slide 11

12 Slide 12

13 Speedup Slide 13

14 Results Analysis Distributed Framework works better than Sequential Execution Expected Speed-Up not achieved  Reading data from the file took more time than expected  Reduced time for the other steps Slide 14

15 Future Work Optimization of reading data from file Use various file systems – NFS/AFS Include more measures of dispersion Increase the number of parameters Slide 15

16 Questions?? Slide 16


Download ppt "Computation Time Analysis - Climate Reanalysis Data Dipanwita Dasgupta University of Notre Dame Graduate Operating Systems."

Similar presentations


Ads by Google