On Tue, 27 May 2008 17:31:38 +0800, "k"
wrote:


http://www.flashmemorytoolkit.com/download.html


the free prog asks for registration code each time it starts but you just
hit 'try' and you're into the full kit

small free and it'll write junk files to your card and read them back to
give you good speed comparisons - don't forget to delete these files
afterwards

I had some *very* surprising results, which supports the many sites out
there that state what manufacturers claim is often not quite true

Many (most?) manufacturer claims are true but are "up to",
maximum figures when the (nnn)X rating versus the more
clearly defined SDHC Class (n) minimum rating.

Flash Memory Toolkit is a handy program but it can't provide
information used to decide the manufactuer claim is untrue.
The manufacturer claim is based on an ideal environment and
transfer, while many of the following can cause a lower
rate:

- Benchmark write access pattern

- Inefficiencies or bottlenecks in the device using the
card.

- Inefficiencies or bottlenecks in the (PC USB) card reader.

- Inefficiencies or bottlenecks in the PC USB itself.

- Inefficiency in the filesystem

- File sizes used for the test, or used in device operation

Consider that many flash memory cards/USB drives/etc using
current technology are rated around 30MB/s read speed, and
yet many PCs can't get 30MB/s in some uses or tests, over
USB, even if a much faster device like a good modern hard
drive was connected via external enclosure instead of a
flash drive. Clearly this means the bottleneck on a good
spec'd flash drive is often elsewhere.