Hi List, First of all, thanks to all who have dedicated time and resources to heavily test Blosc (and most specially to Tony Theodore, who ran the hardsuite on a impressively large variety of OS/processors!). Now, it seems that it runs flawlessly in all the main platforms: Linux, Windows and OSX.
Another interesting outcome of this effort is that you can see the evolution of the machines during the latest decade regarding the increasing mismatch between processor performance and the access to memory: http://blosc.pytables.org/trac/wiki/SyntheticBenchmarks (see at the end of the page) Now, apart of fixing remaining bugs, I've changed the API slightly (added a new parameter to specify a maximum length of the output) and implemented a 0 compression level. The latter may be useful for people that will use Blosc as a general container for its data without compression. Although this mode will not be usable by PyTables directly, it is interesting to see how Blosc can be faster than memcpy() by using this non-compression mode (yes, it seems that memcpy() could get better performance if it would use multi-threading as Blosc does), although it can be worse in others (specially in single-core machines). In order to check that I've introduced no new bugs (I have frozen any further development in Blosc until 1.0 would be out), it would be very interesting if you can participate in the second round of tests. The procedure is very similar as the first round, just go to: http://blosc.pytables.org/trac/wiki/SyntheticBenchmarks and follow instructions there. Ah! by popular demand there is now a win32 binary available (as well as the traditional win64 one). Thanks! -- Francesc Alted ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Pytables-users mailing list Pytables-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/pytables-users
