I am currently working on a C ++ project which includes reading thousands of small (~ 20kb) of text files All are in ASCII
"itemprop =" text ">
string in any number Changing, while the CPU is not cheap in the cycle, it is a non-issue that the amount of overhead associated with I / O is always the order of magnitude larger than the conversion. It is, a disk supplies about 8KB as fast as 20KB, it all comes out of the same cluster on the same track. Including a disk reader head to open a file and it takes forever .
Therefore, focus on reducing the number of files for actual files.
Comments
Post a Comment