Big Chunks vs Small Chunks

Hi, All We just upgraded our data warehouse database(1TB) from 7.31FD2 to 9.40FC2, the instance is running on a partition of HP Superdome attached to EMC storage with 8GB memory cache. There over 500 2GB raw logical devices allocated to the instances with in 66 dbspaces right now which are inherited from ids7 environment before. Since the new ids break 2GB chunk limit to support big chunk, we would like to assess the possibility to convert the large number of small chunks to relative few big chunks. As this is obvious a big move, we would like to listen your opion of such move, such as, . As I understand the new big chunk can reduce the administration of large number devices on Unix level, and possibility of build much bigger database than ids7. What are the else advantage to use big chunk vs small chunk, espacically in terms of performance gain, space overhead etc? . If the advantage of big chunk is un-resistible, what would be best way to do this exercise? Due the size of the database, I can't expect dbimport/dbexport could fullfil this job, I am thinking move table by table with HPL. However, there will require the down time for application. . Please share your experience if you have convert small chunks to large chunks before Thanks in advance. Kent