According to WD’s published specifications, the WD2002FAEX has a maximum sustained data transfer rate of 138 MB/s while the WD5002AALX is rated for 126 MB/s and the WD5003AZEX is rated for 150MB/s.
WD Caviar Black Datasheet:
http://www.wdc.com/wdproducts/library/SpecSheet/ENG/2879-701276.pdf
The slower drive has 500GB per platter technology whereas the WD2002FAEX appears to use short stroked 600GB platters.
Comparing the transfer rates, we have …
(138 / 126)^2 x 500GB = 600GB
A fully stroked drive will have a ratio of 2:1 between the max/min transfer rates, whereas a short stroked 600GB platter would have a ratio that is signficantly less than this, probably 1.7. You should see this on your HD Tune read benchmark graph.
The WD5003AZEX in the datasheet is benchmarking like a drive with a single short stroked 700GB platter:
(150 / 126)^2 x 500GB = 709GB
However, the following thread has a WD5003AZEX model that benchmarks like a fully stroked, 1TB-per-platter drive with a single head:
http://www.tomshardware.com/forum/291056-32-difference-wd5003azex-wd5002aalx-black-500gb
The following example is a bit “fuzzy”, but it appears to show your model with fully stroked 500GB platters:
http://img853.imageshack.us/img853/8307/capturehot.png
As for the Raw Read Error Rate threshold of 200 (I suspect that it is actually 51, not 200), you appear to be confusing normalised values with raw values. Normalised values are “health” scores whereas raw values usually reflect an actual error count. For example, you might find that the normalised value might only drop 1 point (from 200 to 199) when the raw value rises to 1000, say.
BTW, the following statement is quoted from my article:
http://www.users.on.net/~fzabkar/HDD/Seagate_SER_RRER_HEC.html
“The raw values of the RRER attributes represent a sector count, not an error count. This figure rolls over to 0 once the count reaches about 250 million.”
Unfortunately it applies to Seagate’s drives, not WD’s. :-)))
