X-bit labs

Three SD Cards and One MMC Card Reviewed (page 4)

Category: Memory

by Andrey Kuznetcov

[ 07/17/2005 | 10:28 AM ]

Performance in FC-Test

As usual, we resorted to FC-test program to obtain the actual performance results for all the memory cards tested. With the help of X-bit’s own FC-test utility we considered the situations emulated by three major patters, which demonstrated clearly how the file size (1, 10 and 100MB) and the number of files in a set (1, 10 and 100) could affect the performance of the storage media in question.

The first diagram shows how quickly the memory cards allowed saving 100 files 1MB each. Kingston solution from the Elite Pro family is far ahead of the others here. The second fastest would be RiDATA card. The third position among the SD media was taken by another Kingston product. As we have expected, the RS-MMC card was the slowest of all here. Of course, since the technical specifications claimed by the manufacturer could hardly give us any hope for success.

The diagram showing the read speed when we have 100 files, 1MB each, show a slightly different situation. We see that the SD memory cards form a much tighter group this time, although the positions within the group remain unchanged. Kingston SD/256-SCS is the winner, and its slower counterpart is the slowest of the three. RiDATA solution retained its position inbetween, while RS-MMC card is the slowest of all.

When we have a pattern of 10 files 10MB each, all the cards seem to be performing faster during file writing. Although this change is still too insignificant to affect the overall picture.

The diagram built for the reading of 10 files 10MB each again indicates that the performance of all the testing participants tends to grow up a little bit. The general positioning of the racers remained unchanged again.

Once the test file grew up to 100MB in size, the write speed increases. However, this increase in performance again doesn’t change anything generally.

The last diagram in this section shows that there is hardly any performance increase when it comes to reading a single 100MB file. Therefore, the testing participants again retained their positions.