- Sep 18, 2014
- 562
- 45
- 91
So earlier this year i bought a cheap flash drive (8GB) with toshiba 15nm TLC flash and AlcorMicro AU6989SN-GTA flash controller. Curious to see how well it fares with read degradation, i've filled it up with h2wtest test data and ran a speed test. This was the result
After 48 days it looks something like this
Quite a drop there, eh ? It gets better. I did a test afterwards every couple of weeks and speed did indeed drop consistently. But funny thing happed the other day i was testing..
This was after roughly 7.5 months after i first wrote the test data on it. Corruption.
So yeah, this flash drive can't even follow the jedec spec, which supposedly assumes 1 year retention when p/e are exhausted.
If anyone is curious, this were the settings i used for the controller prior of writing the test data.
(there are lots of fun stuff you can do with MPtools and some can affect performance as well).
Anyone has similar findings ?
After 48 days it looks something like this
Quite a drop there, eh ? It gets better. I did a test afterwards every couple of weeks and speed did indeed drop consistently. But funny thing happed the other day i was testing..
This was after roughly 7.5 months after i first wrote the test data on it. Corruption.
So yeah, this flash drive can't even follow the jedec spec, which supposedly assumes 1 year retention when p/e are exhausted.
If anyone is curious, this were the settings i used for the controller prior of writing the test data.
(there are lots of fun stuff you can do with MPtools and some can affect performance as well).
Anyone has similar findings ?