On Thu, 4/24/2025 10:02 AM, Jeff Gaines wrote:
In January I mentioned a problem witch a couple of WD Blue SSDs, purchased in 2020, that had failed.
The replacement for the second one arrived at lunch time today, 2.5 months to get them both replaced.
It has been an interesting exercise, and information from this group suggests that a WD Blue may not have been my best choice.
https://www.newegg.com/western-digital-1tb-blue-sa510/p/N82E16820250229
(Review section)
71% five stars
...
13% one star
"Write Speeds sub-par" 2 stars
"Drive Failed After 3 Months" 1 star
"Worked fine until it died" 2 stars
"Failed after 16 months" 1 star
"Do not buy this drive
until WD fix their QC issue" 1 star
"Absolute trash (20 day lifespan) 1 star
Yes, an interesting set of review samples so far.
Some of their NVMe get better reviews than that.
*******
"1TB Blue sports a 400TBW rating" (sometimes,
a manufacturer doesn't even know how to rate these,
like in the old days)
(A comparable 870 from Samsung, is 600TBW, and
there are devices at the 1TB capacity level
with a higher wear-life-TBW than that.)
TLC chips were supposed to have a "raw" lifespan
of 3000 writes per cell. 600 writes per cell rating
helps codify the write amplification factor as
the controller re-arranges the data storage
during later maintenance (moving from pseudo-SLC
cache to main TLC storage area, or defragging 4K
blocks within a page after a 4K random write test).
The brands of flash, have different internal layouts,
tiling patterns and cell spacing. This can affect
pattern sensitivity (like, when the cells are
getting mushy and close to voltage thresholds). If
two flash brands were "22nm", the center to center
on the cells could be different for the two products.
There's no way for a mere enthusiast review site, to
offer that level of insight about the NAND chips.
Only TechInsights with their electron microscope
and vat of battery acid, can do that sort of analysis
(for a price). There's not really much an enthusiast
site can do, to pre-warn you about expected trouble.
The trouble, after all, could be firmware related,
and the flash itself is "not guilty yer honour".
The firmware in the industry, seemed to go through
a "transition" phase, as if word of how to design
the firmware was making the rounds. It seems
products today, are much better compared to the
"I turned it on and it died" that used to happen
when SSDs first came out.
At first, SSDs had a pad on the PCB for a "Supercap",
which is an energy storage device, and on power-fail,
that provided power until all the cached blocks
in DRAM were flushed. One of the objectives of the
firmware writers, was to *eliminate* the need for
that safety feature. When the power goes off today,
it is the firmware design, that saves your ass.
Paul
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)