which is better for performance named graph within the same dataset or separate datasets

[Abdullah Ibrahim]

Hi everyone,
which is better for performance of sparql queries over named graphs within the 
same the datasets, or make each named graph as separate dataset?

Thanks