Over on the Windows Mobile Team Blog is an excellent post about SD memory cards. I wished I read the post before I bought a SD card. Here’s a brief summary of the post:
SD stands for Secure Digital, but no one ever bother to make it secure. However, the name stuck anyway. SD cards are based on nand flash memory and are used for storing data. SD, MiniSD, and MicroSD (ie – TransFlash) are all the same thing, just in different packaging. Smaller packaging enables smaller phones.
Up to 2G, the major drawback to larger storage cards is they cost more. They don’t burn any more power than smaller ones and generally are not any slower. 4G and larger cards, however, may not work on all devices. Most cards and card readers follow the 1.1 version of the SD specification, which maxs out with 2G cards. The new SD 2.0 standard does allow for larger cards, but some RTOS (such as Windows Mobile) currently doesn’t support SD 2.0. However, it is possible for OEMs to add support for larger SD cards.
SD cards have all sorts of confusing marketing terms like Ultra, 133x, High Speed, etc. However, there’s no standard for what these terms mean, so it’s hard to tell if an “Ultra” card is faster or slower than a “133x” card. There are huge speed differences in different SD cards. A faster card can have a number of benefits. It could show up faster when you plug it in, files could be written to it faster, and files could be read from it faster. Whether or not you’ll be able to actually notice these differences will depend on how much faster the card is and how large the files you’re transferring are. It is also possible to save a bit of power with faster cards. The SD card is only powered while it’s transferring data. If data is transferred quickly, it’ll be powered for less time.
How SD Cards Work
Every SD card has a small microcontroller on it. That microcontroller handles communicating between the device and the card, getting data from the flash, and putting data into the flash. It also handles flash-specific tasks like wear leveling and compaction. The microcontroller only runs while data is being sent to or read from the card. All CPUs and microcontrollers run off of a clock signal, and the SD card’s clock is controlled by the device it is plugged into. When the clock isn’t running, neither is the microcontroller. The device only runs the clock when it wants to transfer data to or from the card.
There are several ways an SD card can be faster than another card. One way one card can be faster than another is that it can have a faster clock speed. The device asks the card how fast a clock it can handle, and then uses that speed. The faster the clock, the more data sent every second. A second way a card can be faster involves the way communications work. When the device issues a command to the card, other bookkeeping operations may also take place. The longer these other operations take, the slower the overall transfer is. Some cards will be more efficient than others, resulting in overall faster transfer times.
More info about SD Memory Cards »