I dont know what you are using cache for exactly, but if if you are transcoding to and/or from /tmp, the difference is likely to be dominated by the extra copying.
ok...copy content to mktmp() dir on tmpfs...do transcode and processing on tmpfs...copy output artifact back to rotational disk...delete tmpfs work dir
experiential data shows a speedup factor of about 4 times when done on tmpfs, even with copying source content to the tmpfs
and the reason to use tmpfs is LESS COPIES, not more. the data is already in main memory and can be mmapped without copying.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.