https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=244356
--- Comment #51 from Olivier Certner <olivier.free...@free.fr> --- (In reply to Hans Petter Selasky from comment #50) Indeed, I observed the 64KiB limit for USB 2.0. I was talking about USB 3.0 sticks, for which I cannot seem to get bigger transactions than 128KiB. Both Linux and MacOSX, according to `iostat`, seem to use 1MiB transactions on these, and seem to get a little bit more bandwidth for sequential access (to a file, not to the disk directly, so FS clustering comes into play, and probably several other things). Unfortunately, I did not keep logs for these tests (on SD_64G and SD_128G, see comment #45), and I don't recall the exact bandwidth advantage. I'm doing a new round of tests with KT_32G, and at least on MacOSX, I get comparable performance (a bit lower in fact) despite the different transaction size. Anyway, the speed differences here are second order to what prompted the opening of this bug, and possibly not even relevant to non-sequential access. -- You are receiving this mail because: You are the assignee for the bug. _______________________________________________ freebsd-usb@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-usb To unsubscribe, send any mail to "freebsd-usb-unsubscr...@freebsd.org"
