next up previous contents
Next: Introduction Up: Platforms for HPJava: Runtime Previous: List of Tables   Contents


List of Figures

  1. Sequential Matrix multiplication in HPJava.
  2. A parallel Matrix multiplication in HPJava.
  3. A general Matrix multiplication in HPJava.
  4. RMI Architecture.
  5. The JMS API Architecture
  6. A parallel matrix addition.
  7. The HPJava Range hierarchy
  8. A general Matrix multiplication in HPJava.
  9. Solution of Laplace equation by Jacobi relaxation.
  10. Example of a distributed array with ghost regions.
  11. Illustration of the effect of executing the writeHalo function.
  12. Principal classes of mpiJava
  13. HPJava data parallel version of the N-body force computation.
  14. Version of the N-body force computation using reduction to Java array.
  15. API of the class BlockMessSchedule
  16. Partial API of the class Range
  17. sendLoop method for Remap
  18. Illustration of sendLoop operation for remap
  19. Jacobi relaxation, re-using communication schedules.
  20. An HPJava communication stack.
  21. The public interface of Buffer class.
  22. The public interface of WriteBuffer class.
  23. The public interface of ReadBuffer class.
  24. The public interface of ObjectWriteBuffer class.
  25. The public interface of ObjectReadBuffer class.
  26. The public interface of mpjdev Comm class.
  27. The public interface of Request class.
  28. Overall layout of logical message.
  29. Layout of primary payload
  30. Layout of one section in the primary payload
  31. Layout of send messages. a) When secondary payload is empty. b) When secondary payload is not empty and the sum of secondary payload size and the size of the primary payload are less than the capacity of the buffer object. c) When secondary payload is not empty and the sum of secondary payload size and the size of the primary payload are greater than the capacity of the buffer object.
  32. Multithreaded implementation.
  33. LAPI Active Message Call.
  34. Layers of a proposed mpjdev reference implementation
  35. Independent clients may find mpjdevService daemons through the Jini lookup service. Each daemon may spawn several slaves.
  36. An example of multigrid iteration.
  37. Red black relaxation on array uf.
  38. Restrict operation.
  39. Illustration of restrict operation
  40. Interpolate operation.
  41. Illustration of interpolate operation
  42. Red-black relaxation of two dimensional Laplace equation with size of $ 512^2$.
  43. Three dimensional Diffusion equation with size of $ 128^3$.
  44. Three dimensional Diffusion equation with size of $ 32^3$.
  45. Multigrid solver with size of $ 512^2$.
  46. A 2 dimensional inviscid flow simulation.
  47. CFD with size of $ 256^2$.
  48. Comparison of the mpjdev communication library using MPI vs. LAPI.
  49. writeHalo() communication patterns on 9 processors.
Abstract

Bryan Carpenter 2004-06-09