You are not logged in.
Pages: 1
I noticed that reading from /dev/urandom won't utilize more than one core, resulting in huge extra time required and wasted CPU potential if you need masses of random data.
Then I saw "urandompar", which makes use of multiple cores,
http://bues.ch/gitweb?p=urandompar.git;a=summary
Anyone know if this is safe? Maybe it could be added to Arch by default to support modern multi-core systems?
Offline
/dev/urandom generates pseudorandom data. If you really need huge amounts of pseudorandom numbers in some userspace application you can use the rand() function from C stdlib or some specialized PRNG library which will likely be faster and/or better than /dev/urandom.
Last edited by mich41 (2013-04-06 11:17:37)
Offline
/dev/urandom generates pseudorandom data. If you really need huge amounts of pseudorandom numbers in some userspace application you can use the rand() function from C stdlib or some specialized PRNG library which will likely be faster and/or better than /dev/urandom.
I'm not writing a C program, but thanks for the suggestion.
Last edited by Jindur (2013-04-06 11:47:54)
Offline
Pages: 1