badblocks sandisk [Solved]

[plopke]

Got a new 480GB sandisk Ultra II , what are the number of badblocks to expect on a new drive?

UPDATE :

Had this question running already at Sandisk who escalated it higher up. They called me and acknowledged what you guys said.

1) It is normal
2) Some firmware is still reporting all production bad blocks , which is not intended.
3) The number of increasing bad blocks is the one to keep a eye on and should not increase by big margins.

 

[quant_123]

I have a new Sandisk x400 1tb with 1122 bad blocks. Hasn't changed after ~322gb writes.

 

[plopke]

I have a new Sandisk x400 1tb with 1122 bad blocks. Hasn't changed after ~322gb writes.

Thx

 

[Billy Tallis]

As another data point, my X400 1TB sample is showing 1182 Total Bad Blocks and 51 Maximum Bad Blocks per Die after 13832 total GB written. I don't have an Ultra II handy to check.

 

[quant_123]

As another data point, my X400 1TB sample is showing 1182 Total Bad Blocks and 51 Maximum Bad Blocks per Die after 13832 total GB written. I don't have an Ultra II handy to check.

Did your x400 start with that amount of bad blocks, like mine. Or did that number accumulate over time?

Out of curiosity, is 13,883gb the amount of total writes (attr 241) or nand writes (attr 233)? Also, what does your percent wear show (attr 230)?

Mine is worn 0.05% after 174gb nand writes and 330gb total writes. That puts write amplification around 0.5, which is very low. A more typical number is between 2-3. My understanding is such a low amplification is possible due to newer LDPC error correction and Sandisk's use of SLC NCache.

 

[Elixer]

That would be odd.
I would have thought they would zero out the SMART table before shipping out?

 

[Mike64]

That would be odd.
I would have thought they would zero out the SMART table before shipping out?

I guess they might have gone that route probably without regulatory complaint, but it would be "falsifying" the drive's specs, no? Like a car dealer zeroing out the odometers on their (new) cars (and/or whatever data is stored in them-thar newfangled electronic sensors cars have these days) cars prior to sale...

 

[Elixer]

I guess they might have gone that route probably without regulatory complaint, but it would be "falsifying" the drive's specs, no? Like a car dealer zeroing out the odometers on their (new) cars (and/or whatever data is stored in them-thar newfangled electronic sensors cars have these days) cars prior to sale...

No, it is actually very common practice in the industry to wipe the SMART stats off the drive once they are done "testing" it.
Lots of HDs & SSDs have bad sectors when they are shipped to the customer, and they remapped them out, and reset SMART.

 

[bigboxes]

I ran the extended test and no bad blocks were reported for my year and a half old Extreme Pro.

 

[quant_123]

I ran the extended test and no bad blocks were reported for my year and a half old Extreme Pro.

SMART shows bad blocks for my drive but does not report them in Sandisk's diagnostics tests. I take this to mean a certain amount of initial bad blocks is normal.

 

[plopke]

Updated and marked it as solved.