CHimera uchime

I want to use uchime for ~600,000 sequences with reference argument as self. When I run the command I get the message that it can be run using only one processor. This is taking very long time.

Is there some way to speed it up? Can I divide sequences into multiple files and use uchime?

Thank you,

Do you only have one group with 600,000 sequences in it? The SOP suggests using the group option which will split the fasta file by groups and then check each group separately. This makes it parallelizable and much much faster (and probably a better analysis).