I work for a government contractor. We supply equipment to them that uses an SSD. This equipment runs 24 hours a day, 7 days a week processing information. It has Win XP Pro so it is not an SSD “aware” OS.
It is my understanding that the average read/write lifespan for each cell is approximately 100K.
My question is this:
Is there a way to see this drive failing BEFORE it actually fails?
An example of this would be since we use a 32GB SSD, when I go to the drives properties it shows up as 29.8GB in Windows. As this drive starts failing because of individual cells going down, will this number change to reflect the current USABLE maximum space?
This would be a very easy way to tell that our SSDs are starting to fail.
What do you think? Is there a way to know before a drive fails?
I strongly doubt disk properties will show you anything, especially if your SSD has spare sectors. And you'd have to check it through the Disk Management, not at the Explorer which, at best, reacts after a Checkdisk.
You should look at the SSD manufacturer's site if it offers supervision software.
Anyway, I don't understand why people bother about Flash write cycles. Thanks to wear levelling, if you write 1GB data per hour, your 100k cycles 32GB disk is expected to last for 3M hours or 300 years. That's huge, it's a wear expectancy and not an Mtbf, and it's much longer than what a mechanical disk does.
if you write 1GB data per hour, your 100k cycles 32GB disk is expected to last for 3M hours or 300 years.
You got that by assuming a write of 32 GB was one "write cycle" for all the cells, then multiplied it by 100K to get 3,000,000 hours, right? There are two serious flaws with your calculation:
1) the 100K write cycles apply to SLC storage chips - most consumer drives use MLC chips which may expire after only 10K write cycles or less
2) because the smallest erasable unit in a flash drive is a "page" that is typically 500KB to 1MB in size, writes often require an erase cycle for a much larger amount of storage than the quantity of data written would suggest. This effect is called "write amplification", and it can make a big difference to the effectiveness of wear leveling, especially in drives that have less sophisticated controllers.
Intel is generally acknowledged to have the best controller in the industry, and they're also the only manufacturer I've seen who actually state the life expectancy of their Solid State Drives: if you write 20GB per day (that's LESS than 1GB per hour), Intel's drives are expected to last for at least 5 years. That's a lot, LOT shorter than your calculation would suggest.
We don't really know, but based on the fact that nobody else is making longevity claims for their SSDs I'd suspect the other drives have a shorter minimum lifespan for equivalent write volumes. That's one of the reasons why people such as myself are willing to pay more for the Intel drives.
1) Right! I own an SLC, as a consequence of some foolish.
2) Disagree about write amplification, because the controller uses Ram and wear levelling to put many small files in a single write block. It will amplify to one sector, not to one write block.
And anyway, a desktop user won't achieve gigabytes with small files. This is possible only with a database and many users.
Replace 100k cycles with 10k for MLC and my calculation drops to 30 years. Write amplification on tiny data from a database may bridge the discrepancy to 5 years, which Intel also calls "at least" and not "a mean value".
Which doesn't change my conclusion: a mechanical disk won't survive as much write. Imagine 20GB a day for 5 years. So don't worry about SSD wear.