Writing by Chunk in PHDF5

In this example each process writes a "chunk" of data to a dataset. The C and Fortran 90 examples result in the same data layout in the file.

Figure a   C Example	Figure b   FORTRAN 90 Example

For this example, four processes are used, and a 4 x 2 chunk is written to the dataset by each process.

To do this, you would:

• Use the block parameter to specify a chunk of size 4 x 2 (or 2 x 4 for Fortran 90).

   

• Use a different offset (start) for each process, based on the chunk size:

  C:
   Process 0        Process 1       Process 2       Process 3
   ---------        ---------       ---------       ---------
   offset[0] = 0    offset[0] = 0   offset[0] = 4   offset[0] = 4 
   offset[1] = 0    offset[1] = 2   offset[1] = 0   offset[1] = 2 
  
  FORTRAN 90:
   offset(1) = 0    offset(1) = 2   offset(1) = 0   offset(1) = 2
   offset(2) = 0    offset(2) = 0   offset(2) = 4   offset(2) = 4
  

For Process 2 the offset and block parameters would look like:

Figure a   C Example	Figure b   FORTRAN 90 Example

Below is an example program for writing hyperslabs by chunk in Parallel HDF5:

C     F90

The following is the output from h5dump for the HDF5 file created with this example:

HDF5 "SDS_chnk.h5" {
GROUP "/" {
   DATASET "IntArray" {
      DATATYPE  H5T_STD_I32BE  
      DATASPACE  SIMPLE { ( 8, 4 ) / ( 8, 4 ) } 
      DATA {
         1, 1, 2, 2,
         1, 1, 2, 2,
         1, 1, 2, 2,
         1, 1, 2, 2,
         3, 3, 4, 4,
         3, 3, 4, 4,
         3, 3, 4, 4,
         3, 3, 4, 4
      } 
   } 
} 
} 

The h5dump utility is a C program, and the output is in C order.