Hi,

When cracking a burst, for all DP generated (with 12lsb == 0), about
30% of them will find a match on disk.
This seems like _a_lot_ !

Especially since when you look at all the end points in the table,
only about 1/750000 of all the possible 52 bits (64-12) values are
present.

This is particularly annoying because :

 - For each hit, we have to replay the chain see if we find the
bitstream we're looking for.
 - If the disk hit rate was very low, we could make an 'index' with
just the endpoints (and this would be about 37% of the current size),
and then just put that part on very fast drive (SSD) and leave the
rest on HDD.


Did anyone notice that / can confirm this ?


Cheers,

    Sylvain
_______________________________________________
A51 mailing list
A51@lists.reflextor.com
http://lists.lists.reflextor.com/cgi-bin/mailman/listinfo/a51

Reply via email to