SD card speed

Intron mag at intron.ac
Mon Aug 14 12:29:28 UTC 2006


Marc van Woerkom wrote:

> Intron wrote:
>> Have you tested your card reader on Microsoft Windows?
>> FreeBSD's FAT module probably isn't optimized enough.
>>
> No, but what I just tried again,after I formatted
> the SD card under Windows.
> 
> Guess what happened:
> 
> tty da2 cpu
> tin tout KB/t tps MB/s us ni sy in id
> 0 64 4.00 483 1.89 10 0 17 13 60
> 0 161 4.00 486 1.90 16 0 13 12 59
> 0 43 4.00 477 1.86 9 0 14 8 70
> 0 65 4.00 422 1.65 8 0 15 9 69
> 0 78 4.00 486 1.90 5 0 11 13 71
> 0 43 4.00 485 1.89 4 0 5 8 83
> 0 121 4.00 474 1.85 10 0 14 9 67
> 0 43 4.00 486 1.90 8 0 16 11 65
> 0 43 4.00 483 1.89 16 0 6 9 69
> 0 43 4.00 396 1.55 7 0 16 9 69
> 0 65 4.00 484 1.89 9 0 14 10 67
> 0 43 4.00 482 1.88 7 0 16 13 63
> 0 43 4.00 485 1.89 5 0 14 12 69
> 0 43 4.00 318 1.24 6 0 9 13 71
> 136 43 4.00 482 1.88 9 0 16 6 69
> 
> Thus the crucial factor is the formatting!
> 

Try to run "chkdsk" against your card under Microsoft Windows XP.
Chkdsk will report file system type (FAT16 or FAT32) and cluster size
(e.g. 16384 bytes) after checking file system on your SD card.

Format your SD card with the same file system type (-F) and cluster size
(-b) under FreeBSD and test the transmission rate again.

As I have referred to in my previous letter, I guess the difference
of transmission rate comes from the asynchronous transmission mechanism
and the sensibility against transmission on-line delay of USB, which
is rather fit for large block transmission than small block transmission.

------------------------------------------------------------------------
                                                From Beijing, China



More information about the freebsd-usb mailing list