1. Enabling Knowledge Discovery in a Virtual Universe Harnessing the Power of Parallel Grid Resources for Astrophysical Data Analysis Jeffrey P. Gardner Andrew Connolly Cameron McBride Pittsburgh Supercomputing Center University of Pittsburgh Carnegie Mellon University

2. How to turn simulation output into scientific knowledge Step 1: Run simulation Step 2: Analyze simulation on workstation Step 3: Extract meaningful scientific knowledge (happy scientist) Using 300 processors: (circa 1995)

3. How to turn simulation output into scientific knowledge Step 1: Run simulation Step 2: Analyze simulation on server (in serial) Step 3: Extract meaningful scientific knowledge (happy scientist) Using 1000 processors: (circa 2000)

4. How to turn simulation output into scientific knowledge Step 1: Run simulation Step 2: Analyze simulation on ??? (unhappy scientist) Using 4000+ processors: (circa 2006) X

5. Mining the Universe can be (Computationally) Expensive The size of simulations is no longer limited by computational power It is limited by the parallelizability of data analysis tools This situation, will only get worse in the future.

6. How to turn simulation output into scientific knowledge Step 1: Run simulation Step 2: Analyze simulation on ??? Using 100,000 processors?: (circa 2012) X By 2012, we will have machines that will have many hundreds of thousands of cores!

7. The Challenge of Data Analysis in a Multiprocessor Universe Parallel programs are difficult to write! Steep learning curve to learn parallel programming Parallel programs are expensive to write! Lengthy development time Parallel world is dominated by simulations: Code is often reused for many years by many people Therefore, you can afford to invest lots of time writing the code. Example: GASOLINE (a cosmology N-body code) Required 10 FTE-years of development

8. The Challenge of Data Analysis in a Multiprocessor Universe Data Analysis does not work this way: Rapidly changing scientific inqueries Less code reuse Simulation groups do not even write their analysis code in parallel! Data Mining paradigm mandates rapid software development!

9. How to turn observational data into scientific knowledge Step 1: Collect data Step 2: Analyze data on workstation Step 3: Extract meaningful scientific knowledge (happy astronomer)

10. The Era of Massive Sky Surveys Paradigm shift in astronomy: Sky Surveys Available data is growing at a much faster rate than computational power.

11. Good News for “Data Parallel” Operations Data Parallel (or “Embarrassingly Parallel”): Example: 1,000,000 QSO spectra Each spectrum takes ~1 hour to reduce Each spectrum is computationally independent from the others There are many workflow management tools that will distribute your computations across many machines.

12. Tightly-Coupled Parallelism (what this talk is about) Data and computational domains overlap Computational elements must communicate with one another Examples: Group finding N-Point correlation functions New object classification Density estimation

13. The Challenge of Astrophysics Data Analysis in a Multiprocessor Universe Build a library that is: Sophisticated enough to take care of all of the nasty parallel bits for you. Flexible enough to be used for your own particular astrophysics data analysis application. Scalable: scales well to thousands of processors.

14. The Challenge of Astrophysics Data Analysis in a Multiprocessor Universe Astrophysics uses dynamic, irregular data structures: Astronomy deals with point-like data in an N-dimensional parameter space Most efficient methods on these kind of data use space- partitioning trees. The most common data structure is a kd-tree.

15. Challenges for scalable parallel application development: Things that make parallel programs difficult to write Thread orchestration Data management Things that inhibit scalability: Granularity (synchronization) Load balancing Data locality

16. Overview of existing paradigms: GSA There are existing globally shared address space (GSA) compilers and libraries: Co-Array Fortran UPC ZPL Global Arrays The Good: These are quite simple to use. The Good: Can manage data locality well. The Bad: Existing GSA approaches tend not to scale very well because of fine granularity. The Ugly: None of these support irregular data structures.

17. Overview of existing paradigms: GSA There are other GSA approaches that do lend themselves to irregular data structures: e.g. Linda (tuple-space) The Good: Almost universally flexible The Bad: These tend not to scale even worse than the previous GSA approaches. Granularity is too fine

18. Challenges for scalable parallel application development: Things that make parallel programs difficult to write Thread orchestration Data management Things that inhibit scalability: Granularity Load balancing Data locality GSA

19. Overview of existing paradigms: RMI rmi_broadcast(…, (*myFunction)); RMI layer myFunction() RMI Layer myFunction() RMI Layer myFunction() RMI Layer myFunction() Proc. 0 Proc. 1Proc. 2 Proc. 3 Master Thread Computational Agenda myFunction() is coarsely grained “Remote Method Invocation”

20. Challenges for scalable parallel application development: Things that make parallel programs difficult to write Thread orchestration Data management Things that inhibit scalability: Granularity Load balancing Data locality RMI

21. N tropy: A Library for Rapid Development of kd-tree Applications No existing paradigm gives us everything we need. Can we combine existing paradigms beneath a simple, yet flexible API?

22. N tropy: A Library for Rapid Development of kd-tree Applications Use RMI for orchestration Use GSA for data management

23. A Simple N tropy Example: N-body Gravity Calculation Cosmological “N-Body” simulation 100,000,000 particles 1 TB of RAM 100 million light years Proc 0Proc 1Proc 2 Proc 5Proc 4Proc 3 Proc 6Proc 7Proc 8

24. A Simple N tropy Example: N-body Gravity Calculation ntropy_Dynamic(…, (*myGravityFunc)); N tropy master layer N tropy thread service layer myGravityFunc() N tropy thread service layer myGravityFunc() N tropy thread service layer myGravityFunc() N tropy thread service layer myGravityFunc() Proc. 0 Proc. 1Proc. 2 Proc. 3 Master Thread Computational Agenda Particles on which to calculate gravitational force P1…P2Pn…

25. A Simple N tropy Example: N-body Gravity Calculation Cosmological “N-Body” simulation 100,000,000 particles 1 TB of RAM 100 million light years To resolve the gravitational force on any single particle requires the entire dataset Proc 0Proc 1Proc 2 Proc 5Proc 4Proc 3 Proc 6Proc 7Proc 8

26. A Simple N tropy Example: N-body Gravity Calculation N tropy thread service layer myGravityFunc() N tropy thread service layer myGravityFunc() N tropy thread service layer myGravityFunc() N tropy thread service layer myGravityFunc() Proc. 0 Proc. 1Proc. 2 Proc. 3 N tropy GSA layer 0 43 1 2 76 5 1110 8 9 1413 12

27. N tropy Performance Features GSA allows performance features to be provided “under the hood”: Interprocessor data caching < 1 in 100,000 off-PE requests actually result in communication. RMI allows further performance features Dynamic load balacing Workload can be dynamically reallocated as computation progresses.

28. N tropy Performance 10 million particles Spatial 3-Point 3->4 Mpc No interprocessor data cache, No load balancing Interprocessor data cache, No load balancing Interprocessor data cache, Load balancing

29. Why does the data cache make such a huge difference? myGravityFunc() Proc. 0 0 43 1 2 76 5 1110 8 9 1413 12

30. N tropy “Meaningful” Benchmarks The purpose of this library is to minimize development time! Development time for: 1. Parallel N-point correlation function calculator 2 years -> 3 months 2. Parallel Friends-of-Friends group finder 8 months -> 3 weeks

31. Conclusions Most approaches for parallel application development rely on a single paradigm Inhibits scalability Inhibits generality Almost all current HPC programs are written in MPI (“paradigm-less”): MPI is a “lowest common denominator” upon which any paradigm can be imposed.

32. Conclusions Many “real-world” problems, especially those involving irregular data structures, demand a combination of paradigms N tropy provides: Remote Method Invocation (RMI) Globally Share Addressing (GSA)

33. Conclusions Tools that selectively deploy several parallel paradigms (rather than just one) may be what are needed to parallelize applications that use irregular/adaptive/dynamic data structures. More Information: Go to Wikipedia and seach “Ntropy”