Requesting Write-Speed Tests of SD Cards in USB 2.0 Readers (some results posted)

BACKGROUND
I am a bit curious as to whether so-called "high speed" SD cards are really that much faster than regular SD.

What sparked my interest in this is a test I did of a 160MB-write on my SimpleTech 512MB SD card in a USB 2.0 reader. The card repeatedly achieved a real-world write speed of about 26.6X or 3.9MB/sec.

The card is not labeled as high speed, and there is no "##X" label on it. ("1X" stands for 150KB/sec.) It appears to be a "normal" card.

However, many memory cards are labeled as 32X in order to distinguish them as capable of "high speed" reading/writing. (Writing is almost always slower, and is usually the speed described by the X-number.) This is also almost always just a "possible" speed, i.e., the card will write at speeds "up to" the one advertised, but real-world speed will be somewhat slower.

One might surmise that if a 32X Lexar SD card attains "real-world" write speeds of, say, 26X (3.8MB/sec), or 81% of its maximum possible speed, then the 32X high-speed label might actually be misleading because my SimpleTech "regular" SD card actually writes faster than that.

TESTS REQUESTED
You will need:
- appropriate SD card (see list below)
- USB 2.0 reader
- USB 2.0-enabled PC
- stopwatch (most digital watches have this feature. i suppose you can use a benchmark program if you have one, but it needs to be trustworthy)
- large-MB files (i.e. videos)
- about 7 minutes to run the test(s) and post the results here (this is a total time estimate, i.e., it should take you about 7 minutes to run the tests, crunch the 4 or 5 numbers that need crunching, type the results up on aximsite, and click "submit post." The tests themselves do not take 7 minutes each.)

Priority SD cards--the ones I am most interested in:
- Regular Sandisk
- Regular Lexar
- Regular SimpleTech
- Regular Kingston
- Sandisk Ultra
- Sandisk Ultra II
- 32X Lexar
- SimpleTech ProX
- any other SD cards described as "high speed"

Priority Capacities (all are welcome):
- 256MB
- 512MB


PROCEDURE
It's quite simple, really:
1. Put memory card in PC card reader. Keep running programs to a minimum. This is not a Pocket PC test!
2. Cut and paste the current contents into your computer before running the test.
3. Find a single file for use as a test file. It should be larger than 50% of the capacity of the card. *If you are testing a 256MB or 512MB SD card, please use a file around 160MB, since that's what I used in my tests.
4. Paste the test file onto the SD card, and start your stopwatch as soon as you see the transfer begin.
5. Stop the stopwatch when the transfer ends.
Repeat as necessary for each card. I am requesting write tests only. Run each test twice if possible...

Result calculation:
There are two numbers I am interested in, the MB transfer rate and the real-world X-rating.
1. Find the total number of seconds elasped in the transfer (...multiply the number of minutes by 60 and add this to the seconds figure....I will call this number "t").
2. Take the file size in MB (if you only have the size in KB, take that number and divide by 1,024 to get the MB size, or simply move your mouse over the file; size in MB should appear) and divide this by t:
MB/t = [MB transfer rate]
3. To find the real-world "X" rating, take the MB transfer rate and multiply by 1,024 to get KBs, then divide by 150 to get the X rating:
[MB transfer rate] *1024 = [KB transfer rate]
[KB transfer rate] / 150 = [X-rating for write speed]
4. (optional) If you know what the card's advertised speed was, find the percentage of that speed achieved by dividing either your MB transfer rate or your X-rating by the rate or rating described by the advertised speed, then multiply by 100. Be sure to match your units...
test result / advertised speed *100 = percentage of advertised speed achieved


RESULT REPORTING
After running your test(s), please provide the following 8 items:
1- Card capacity and brand (specify SD or CF etc., in case some CF tests are posted)
2- Special name, if applicable (like "Ultra II"); if none, type "regular"
3- Advertised write speed information; indicate whether max or min
4- Approximate age of card (time since original purchase date)
5- Test file size used
6- Test Result - MB transfer rate (in MB's per second)
7- Test Result - X-rating (a number followed by an X)
8- Test Result (optional) - percentage of advertised speed achieved

Results will inevitably be skewed by the type of system conditions used for the testing, especially the variation in card readers. Hopefully everything will be "standard" enough for judicious comparison.

Thanks!!!!!!!


PERSONAL RESULTS AND FURTHER DISCUSSION:
1- 512 SimpleTech SD
2- regular
3- up to 10MB/sec
4- 9 months old
5- 160MB
6- 3.9MB/sec
7- 26.6X
8- 39%

1- 64 Viking SD
2- regular
3- assume up to 10X or 1.46MB/sec
4- 12 months old
5- 39.3MB
6- 1MB/sec
7- 6.9X
8- 69%

1- 32 Kingston SD
2- regular
3- assume up to 10X or 1.46MB/sec
4- 3 months old
5- 26MB
6- 1MB/sec
7- 6.9X
8- 69%

1- 1GB SimpleTech CF
2- ProX
3- up to 4MB/sec or 27.3X
4- 1 month old(?)
5- 160MB
6- 2.2MB/sec
7- 15X
8- 55%

1- 512 Viking CF
2- regular
3- assume up to 10X or 1.46MB/sec
4- 1 week old
5- 160MB
6- 1.51MB/sec
7- 10.3X
8- 103%

1- 256 Lexar CF
2- 40X
3- up to 40X
4- 1 week old
5- 160MB
6- 2.9MB/sec
7- 20X
8- 50%

In absolute terms, the SimpleTech SD card is the fastest of this sample (3 SD cards, 3 CF cards) at 3.9MB/sec. Two of the SD cards were very low capacity (meaning old and not designed for speed) so it was not a fair race.

Unless there is something mysteriously biased about my internal 6-in-1 USB 2.0 card reader, these results appear to disagree with the conventional wisdom that CF is usually faster than SD memory cards. The SimpleTech 512MB SD card turned out to be 33% faster than the fastest CF card (Lexar 256MB 40X)tested.

What makes me really curious is this new "ProX" standard that SimpleTech introduced. it's really just their way of saying high speed, but I am suspicious about it because the advertised write speeds for ProX are the same as the old write speeds for regular 256 and 512 SD (neither denomination is listed under regular SD any longer, only under ProX).

I suspect SimpleTech made no significant changes to their 256 and 512MB SD cards except for changing the name in order to make them more appealing and competitive.

This can only be determined by testing to see whether ProX SD cards achieve real-world write speeds significantly higher than 3.9MB/sec, as demonstrated by my "normal" SimpleTech SD card.

And of course, I am curious as to which brand has the fastest average speed, and whether there are some brands who have been shirking excessively on their write speed promises.

Thank you for your assistance!

more info on SimpleTech flash memory cards:
http://www.simpletech.com/flash/index.php

bump

oh no I can't post anymore!!!!!! as of this post, I am 1337!!!!!!!!! help! can admin freeze my post count at this number???? pleeease

l33t!

ok, i respected the l33t gods as long as I could, the screenshot will have to do.

anyway, hope you all dont mind if i keep bumping this until somebody with all the necessary tech sees it and helps me out....

[sheepy]

I've got a 256MB non-32x Lexar, I'll give that a tryout for you this evening :)

mark

[sheepy]

OK, here goes...

1- Card capacity and brand: Lexar 256MB (made in Japan)
2- Special name: Regular
3- Advertised write speed information: not specified on the card
4- Approximate age of card: 4.5 months
5- Test file size used: 165,134KB - 161.26MB

Run 1:
6- MB transfer rate: 2.54MB/sec
7- X-rating: 17.34x

Run 2:
6- MB transfer rate: 1.55MB/sec
7- X-rating: 10.58x

Run 3:
6- MB transfer rate: 1.94MB/sec
7- X-rating: 13.26x

Run 4:
6- MB transfer rate: 1.99MB/sec
7- X-rating: 13.59x

I'm at a little bit of a loss to explain the huge variations in the speeds, the second one went especially slowly (hence why i did another couple of runs...).

hmm, very interesting....

thanks for running the tests! and you were also one of the people who ran battery tests when I requested those....double thanks!

This is useful for me since I'm quite certain I have an identical card (Lexar made in Japan), though it's at home right now (versus here at college) so I can't test it.

Does it look like this?

[sheepy]

The front looks like that, but the back is different... the writing is rotated 90 degrees clockwise and shifted to the top (bottom? opposite end of the sd to the contacts)... but yeah, that's the one.

I'd be interested to see that one alongside the 32x to see what the difference actually is.

Originally posted by sheepy
The front looks like that, but the back is different... the writing is rotated 90 degrees clockwise and shifted to the top (bottom? opposite end of the sd to the contacts)... but yeah, that's the one.

ah yes, i used to have one of those.

I suspect a 32X card would be identical except for a red circle saying "32X" on the front label. However, 32X Lexars were known to be made in China or Taiwan in the past (about 12ish months ago), by Sandisk's OEM, no less. This may not be the case at all today, of course.

I really wonder what a Sandisk SD looks like close-up, cuz I know what a Toshiba OEM SD card looks like, and it's definite and distinct from Panasonic/other Japan OEM cards (such as the one pictured above). In retrospect, all of the Kingston SDs I have ever seen (about 7 or 8 of them) plus at least one PNY and one Viking were all made--or at least assembled--by Toshiba.

Toshiba cards have a distinct shell. The plastic is different from that used by panasonic; it has slightly less sheen to it. also, the write-protection switch is yellow, and finally, the face of the card is recessed in the bottom half where Toshiba intended that a label be attached. Our lexars clearly have the whole front face recessed with the expectation of a full-face label.

I have yet to see an SD card with black plastic, as are often shown in online store pictures...

also i just remembered that my simpletech 512 SD is exactly like your Lexar 256--serial number info at the bottom of the back of the card--except of course it is 512MB and the label says SimpleTech. However, the cards are very nearly identical superficially.

below are some examples of cards made by Toshiba (in both Taiwan and Japan...I have never personally seen a China card).

[Odie2004]

I'll try and post the results with my "high speed" 256mb Panasonic SD card later on today.

- Odie

1- Card capacity and brand: 256 SMART SD
2- non-high speed
3- assume expected write speed = 10X or ~1.46MB/sec
4- Approximate age of card: several days old
5- Test file size used: 160MB
6- Test Result - MB transfer rate: 1.45MB/sec
7- Test Result - X-rating: 9.93X
8- Test Result - percentage of advertised speed achieved: 99.3%

[Odie2004]

1- Card capacity and brand - 256mb Panasonic (made in Japan)
2- "High Speed" marking on card packaging
3- 10 MB/s (I'm pretty sure that this is the Max read speed, the packaging and documentation don't state)
4- Approx. 3 weeks old
5- Test file size used: 160 MB
6- Run 1: 2.54 MB /s
Run 2: 2.54 MB /s
Run 3: 2.5 MB /s
7- Test Result - 17.3 X

thanks for running the test! that is verrrrry interesting result, as it has been surmised that SimpleTech and other Japan cards are actually made by Panasonic. The discrepancy between that card and my 512's write speed seem to suggest otherwise, but it'd be good to get a test of a SimpleTech 256 SD to be sure.

I wonder if the differences between our test system (i.e. computers and readers) is making a big difference here....

[Odie2004]

Originally posted by Astounded EEG
I wonder if the differences between our test system (i.e. computers and readers) is making a big difference here....

It's definately possible. I'm running a relatively slow rig (1.7 P4 (A). It's been running quite sluggishly lately)

I'm using a PQI TravelFlash USB 2.0 card reader. It's pretty small and compact (not to mention cheap!).

- Odie

you call that slow!?!? :) I was using a 500mHz Celeron until last august!

my current system is somewhat improved :). 2.4 GHz Celeron, 512MB DDR, and USB 2.0 of course.

I use an internal 6-in-1 USB 2.0 reader by a no-name brand, but my SimpleTech Bonzai portable 2.0 reader confrims the results i get with the internal, so I reckon it's trustworthy.

some of the tests were done when my system seemed a little slow, but when I ran them again after a reboot, I got the same results, so I don't think system slow downs affect USB transfer speeds too much.


...and i'm rather struck by the fact that most of the people who have assisted me in testing have all come from the UK. Odie on these SD card tests, sheepy on both the SD cards and some X3i battery tests I asked for previously, and happycheesecake helped me out by running some battery tests of his h2210 over at ipaqhq.com.

.....hmm....

[Odie2004]

lol, maybe people from the UK are just naturally more helpful than people from *cough*other*cough* countries ;)