From Newsgroup: alt.windows7.general

On 2025/9/21 4:0:59, Paul wrote:

[]

Purely for your amusement, you could try a bad block scan of the drive.
This is pretty old software itself, having been released in 2008. The last column over will check whether the blocks have bad CRC. That version of HDTune, does not do writes, so it should not make the SSD health any worse.

https://www.hdtune.com/files/hdtune_255.exe

Does that mean _anything_ that that version of HDTune does is safe to do
on an SSD, or only the last column (tab, "Error Scan")? (Doing that now
on my drive - which CrystalDiskInfo says is "CT480BX500SSD1 : 480.1 GB",
and has always said it is "Good, 97%". For interest, it took just under
18 minutes to pass 300000 MB [all green so far]. Finished after 27:10,
457679 MB, all green.)>

There are a couple possibilities. The softwares might not know how to read the SMART table on that particular device. But just as easily, you could already

Interesting: Boris's screenshots show: for HDTune's Info tab, the Failed attribute as "(EE) (unknown attribute)" (when I select that tab, I get a
blank table - how do I get it to populate?), and for CrystalDiskInfo,
the red one as "EE Media Wearout Indicator (Cycles Remaining)", between
C7 and F1 - my CDI doesn't include EE.

[]

The prices on this drive (870EVO), are not always that good. There was a time when Samsung was selling them at a more competitive rate, but that is
not today. I think I have about five of those, all in good working shape.

(Presumably the generally favourable PCWorld review had at least some
effect on the price.)

They're 600TBW per a 1TB drive. A 4TB drive in that family would be 2400TBW. Each cell can be written 600 times. For comparison, I have eight NS100 drives,
which are not in the same class, but are great for doing OS installs and
for testing. Some of them were struggling to perform, on the date of delivery. But that is what you expect when the manufacturer can swap NAND chips in mid production and use something else. It's up to you to
pick something that isn't too sketchy. You can "do sketchy" if you can convince the operator of the computer to "do frequent backups". If I had personal data on the NS100 drives, I would be backing them up for sure.
But they sure are cheap :-)

https://www.pcworld.com/article/407542/best-ssds.html

https://www.pcworld.com/article/393933/samsung-870-evo-sata-ssd-review-the-speed-you-need-at-sane-prices.html

Presumably an SSD sitting unpowered on the shelf isn't deteriorating
fast, other than the guarantee. I just wonder if I should be buying one
in case they cease to be available in that form factor (2.5" drive
equivalent), and/or a capacity that Windows (7 and) 10 can handle (2GB
for 7-32 is it? I don't know for 10). Doesn't look like they're
disappearing yet, though, and the guarantee consideration mitigates
against purchase, at least while my current one stays at 97%?>

Some computers equipped with SATA SSDs, the Southbridge SATA port only ran at SATA II rates.
This also means the power consumption of a SATA III drive running on a SATA II port,
is reduced, as the drive is never running at max speed. While the 870EVO
has a claimed SLC cache, I can't say I've seen SLC cache behavior on it.
Even large writes seem consistent. And the OS decides "how fast it is on small files". A 4K synthetic test can show how fast a drive is, but the
OS can degrade the living shit out of a fine performance there.

(Crystal says mine is using "Transfer Mode: SATA/600 | SATA/600".)>

Anyway, just for fun, run the bad block scan in the last column of HDTune ("Error Scan").
The bad blocks located, are not stored in $BADCLUS, so there is no permanent record
of the mess it finds. You don't want anything compromising the drive contents,
any more than they are already. If you get a lot of "red blocks" in the scan, this means the drive ran out of spare blocks, and could not repair the blocks it was finding as bad ones. The "red blocks" then, are an indicator of how far into debt the drive has gone. Finding red blocks, means that some
of the more simple-minded copy schemes, are going to have trouble.
It then depends on your expectations on data recovery, exactly how
much you can shovel off there. ddrescue (gddrescue package, linux) is
one way to do a low level copy of something in serious trouble. If the scan is all green blocks, then you are laughing. Sure, some of the files are corrupted
(by the inability of the flash to do a good write, or by bit rot), but if the drive really has all green blocks, you can use Macrium or anything else you might like to use. When the red blocks show up, the recovery will be a
battle of wills, you versus the drive :-) Been there, the drive has beaten
me every time.

Paul

--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

Sometimes I'm so sweet even I can't stand it. ~ Julie Andrews
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.windows7.general

On Sun, 9/21/2025 4:23 AM, J. P. Gilliver wrote:

On 2025/9/21 4:0:59, Paul wrote:

[]

Purely for your amusement, you could try a bad block scan of the drive.
This is pretty old software itself, having been released in 2008. The last >> column over will check whether the blocks have bad CRC. That version of
HDTune, does not do writes, so it should not make the SSD health any worse. >>
https://www.hdtune.com/files/hdtune_255.exe

Does that mean _anything_ that that version of HDTune does is safe to do
on an SSD, or only the last column (tab, "Error Scan")? (Doing that now
on my drive - which CrystalDiskInfo says is "CT480BX500SSD1 : 480.1 GB",
and has always said it is "Good, 97%". For interest, it took just under
18 minutes to pass 300000 MB [all green so far]. Finished after 27:10,
457679 MB, all green.)>

There are a couple possibilities. The softwares might not know how to read >> the SMART table on that particular device. But just as easily, you could already

Interesting: Boris's screenshots show: for HDTune's Info tab, the Failed attribute as "(EE) (unknown attribute)" (when I select that tab, I get a blank table - how do I get it to populate?), and for CrystalDiskInfo,
the red one as "EE Media Wearout Indicator (Cycles Remaining)", between
C7 and F1 - my CDI doesn't include EE.

[]

The prices on this drive (870EVO), are not always that good. There was a time
when Samsung was selling them at a more competitive rate, but that is
not today. I think I have about five of those, all in good working shape.

(Presumably the generally favourable PCWorld review had at least some
effect on the price.)

They're 600TBW per a 1TB drive. A 4TB drive in that family would be 2400TBW. >> Each cell can be written 600 times. For comparison, I have eight NS100 drives,
which are not in the same class, but are great for doing OS installs and
for testing. Some of them were struggling to perform, on the date of
delivery. But that is what you expect when the manufacturer can swap NAND
chips in mid production and use something else. It's up to you to
pick something that isn't too sketchy. You can "do sketchy" if you can
convince the operator of the computer to "do frequent backups". If I had
personal data on the NS100 drives, I would be backing them up for sure.
But they sure are cheap :-)

https://www.pcworld.com/article/407542/best-ssds.html

https://www.pcworld.com/article/393933/samsung-870-evo-sata-ssd-review-the-speed-you-need-at-sane-prices.html

Presumably an SSD sitting unpowered on the shelf isn't deteriorating
fast, other than the guarantee. I just wonder if I should be buying one
in case they cease to be available in that form factor (2.5" drive equivalent), and/or a capacity that Windows (7 and) 10 can handle (2GB
for 7-32 is it? I don't know for 10). Doesn't look like they're
disappearing yet, though, and the guarantee consideration mitigates
against purchase, at least while my current one stays at 97%?>

Some computers equipped with SATA SSDs, the Southbridge SATA port only ran at SATA II rates.
This also means the power consumption of a SATA III drive running on a SATA II port,
is reduced, as the drive is never running at max speed. While the 870EVO
has a claimed SLC cache, I can't say I've seen SLC cache behavior on it.
Even large writes seem consistent. And the OS decides "how fast it is on
small files". A 4K synthetic test can show how fast a drive is, but the
OS can degrade the living shit out of a fine performance there.

(Crystal says mine is using "Transfer Mode: SATA/600 | SATA/600".)>

Anyway, just for fun, run the bad block scan in the last column of HDTune ("Error Scan").
The bad blocks located, are not stored in $BADCLUS, so there is no permanent record
of the mess it finds. You don't want anything compromising the drive contents,
any more than they are already. If you get a lot of "red blocks" in the scan,
this means the drive ran out of spare blocks, and could not repair the blocks
it was finding as bad ones. The "red blocks" then, are an indicator of how >> far into debt the drive has gone. Finding red blocks, means that some
of the more simple-minded copy schemes, are going to have trouble.
It then depends on your expectations on data recovery, exactly how
much you can shovel off there. ddrescue (gddrescue package, linux) is
one way to do a low level copy of something in serious trouble. If the scan >> is all green blocks, then you are laughing. Sure, some of the files are corrupted
(by the inability of the flash to do a good write, or by bit rot), but if the
drive really has all green blocks, you can use Macrium or anything else you >> might like to use. When the red blocks show up, the recovery will be a
battle of wills, you versus the drive :-) Been there, the drive has beaten >> me every time.

HDTune 2.55 does not have the SSD SMART table information needed, to do a
good job of decoding SSD SMART. However, when the table won't render at all, your drive could be in a USB enclosure where the SMART Passthru option is
not present. And storage devices which don't have SMART, will also not
display a SMART table. And my 4GB HDD from the year 2000, does not have
SMART either :-) That means it is perfectly healthy. But if SMART is accessible, the (dumb) HDTune will try to decode it anyway, with comical results. If any tables look goofy in HDTune, don't take it personally.
Whereas CrystalDiskInfo is more carefully managed, and a lot more drives
and their quirks, should be handled properly.

You would use something like CrystalDiskInfo, to display a wider range of
SMART tables. And when an SSD is branded, like an 870EVO from Samsung,
you can go look for a "Samsung ToolBox" (each company has a peppy name
for their software) and install that.

[Picture]

https://i.postimg.cc/L4fQ0qpB/toolkit-99-percent-life-left.png

# Samsung Magician

https://semiconductor.samsung.com/consumer-storage/support/tools/

While the SSDs have some common entries, you can find vendor-specific
entries on a lot of them. And we don't know what the vendor-specific
entries mean. It's so bad in fact, that some drives, the Toolkit
does not support several of the drive models. They are that non-standard
and unsupported. That means the drive manufacturer (!) cannot display
the SMART, and quite possibly, the CrystalDiskInfo designer doesn't
know the answer either. The drive manufacturer likely does know,
but the info may be covered by an NDA preventing the sharing of the
info with mere customers.

HDTune 2.55 is perfectly safe as far as I know. If any info displayed
is not to your liking, the software is quite old and a lot of things
have happened since that free version was frozen. The supported (paid)
version is much nicer looking and supports R/W characterization for
things like cheesy SMR (shingled HDD) drives. The read and write speeds
do not need to be identical on HDD, and that would happen on an SSD only
if the error corrector ARM cores can't keep up with the TLC read errors :-)

TLC or QLC drives can get "mushy" if left unpowered for periods of time.
For drives that are powered irregularly but at some frequency, the
SSD drives like that have maintenance routines they can run to freshen
the storage (at the cost of one cell wear for the affected areas).
The "uncorrected" error information, helps the drive figure out how
mushy things are. For example, sectors in the free pool, would not
need to be "tended", as they don't contain user info. So if I had
a 4TB drive with 1TB of partitions, only 1TB of it would be a candidate
for "freshening". Windows 10 running TRIM once a week, helps inform the
drive about the status. You can create a temporary partition on an
unallocated part of the drive, TRIM the partition, then delete the
partition, as a way of signaling an unallocated section is not being
used. The drive does not "know" the partition structure -- the TRIM
just gives it a list of LBAs known not to be in usage, and the calculation
is based on partition file contents and their used clusters.

They're trying to make SATA disappear (the 870E boards only have four
SATA ports, versus the traditional six SATA ports). I think it is a
reasonably estimate, that "some day", SATA will disappear just as
surely as IDE. But there is really no need of doing that. SATA is
a pretty economical standard, and they should always be able to support it.

You should look at the prices of NAND, predictions of the future,
before buying a stocking stuffer for later. At some previous time,
I was warning people that at the time, that was a PERFECT TIME
to buy an SSD. But that time has passed and the price is a bit high
at the moment. If you find drives at lesser prices (like the NS100 drives
I used for scratch OS installs), those are not recommended for long
term storage, just because of how the latest ones have presented themselves when brand new. The first batch I bought, were much better behaved
than the current ones. And since there is a Chinese NAND flash company
now, it is always possible some of those will become "favoured"
for usage in super-cheap drives ($40). There are two aspects to
NAND, knowing how to make them, but also screening them and deciding
who gets the table scraps. We can see the results of table-scrap NAND
on the USB sticks we buy. We don't want that grade of NAND in our
SSD drives.

Presumably, the NAND companies characterize their NAND layers, before
they stack them. There are two layers of layers. There is the vertical
string in a NAND slab. And you can stack NAND slabs to make a sandwich.
That's how you squeeze 2TB into a single chip, is stacking slabs.
Without slabs, it would be harder to make 256GB drives (with a single
256GB chip), and 4TB drives (with two 2TB chips). And you would want
to screen the slabs before stacking, to cut down on material loss.
In the same way as MCM (multi-chip module packaging), requires
the usage of screened die, to keep the yield of MCM packages up.

https://www.atpinc.com/blog/what-is-nand-die-stacking

You can even stack DRAM, but fortunately the practice is not common yet (thermals, when doing Refresh).

The GPT issue with Windows 7, might have been with respect to booting
or something. I would not particularly expect a limitation on
only being able to use 2TB drives on Windows 7 x86. If you wanted
to use a 4TB GPT drive for data on Windows 7 x86 that should work.
The Wiki has the detail on booting.

https://en.wikipedia.org/wiki/GUID_Partition_Table

"Details of GPT support on 32-bit editions of Microsoft Windows[33]"

And when cloning a drive, I think you can see the potential for trouble,
when taking a <2TB drive content and placing it on a >2TB device.
The former drive may be MBR partitioned, the new drive may need to be
GPT, yet the cloning softwares may not be equipped for the "transition".
And partition manager softwares, may not have good coverage for this
case either. While Windows has an "MBR2GPT.exe" utility, that is
mainly intended for boot drive transitions and not for general
purpose partition handling. I've not tested that in a VM to see
what kind of a mess it makes of data drives. That's a research topic
for somebody.

Paul

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.windows7.general

On 2025/9/21 13:50:20, Paul wrote:

On Sun, 9/21/2025 4:23 AM, J. P. Gilliver wrote:

On 2025/9/21 4:0:59, Paul wrote:

[]

Purely for your amusement, you could try a bad block scan of the drive.
This is pretty old software itself, having been released in 2008. The last >>> column over will check whether the blocks have bad CRC. That version of
HDTune, does not do writes, so it should not make the SSD health any worse. >>>
https://www.hdtune.com/files/hdtune_255.exe

Does that mean _anything_ that that version of HDTune does is safe to do
on an SSD, or only the last column (tab, "Error Scan")? (Doing that now
on my drive - which CrystalDiskInfo says is "CT480BX500SSD1 : 480.1 GB",

[]

HDTune 2.55 does not have the SSD SMART table information needed, to do a good job of decoding SSD SMART. However, when the table won't render at all, your drive could be in a USB enclosure where the SMART Passthru option is

No, it's where the HD was in this Lenovo ideapad. (At least I assume it
is; I bought the ideapad with it already fitted by the refurbisher [a
small shop]).

CrystalDiskInfo read its SMART stull no problem.

not present. And storage devices which don't have SMART, will also not display a SMART table. And my 4GB HDD from the year 2000, does not have
SMART either :-) That means it is perfectly healthy. But if SMART is accessible, the (dumb) HDTune will try to decode it anyway, with comical results. If any tables look goofy in HDTune, don't take it personally.

It just showed blank - column headings, with nothing below them. no
problem - I'll just use Crystal. I hadn't even installed HDTune until
you said (to Boris) it was safe to use the last tab (to look for red
blocks).

Whereas CrystalDiskInfo is more carefully managed, and a lot more drives
and their quirks, should be handled properly.

You would use something like CrystalDiskInfo, to display a wider range of SMART tables. And when an SSD is branded, like an 870EVO from Samsung,
you can go look for a "Samsung ToolBox" (each company has a peppy name
for their software) and install that.

[Picture]

https://i.postimg.cc/L4fQ0qpB/toolkit-99-percent-life-left.png

# Samsung Magician

https://semiconductor.samsung.com/consumer-storage/support/tools/

Crystal just says it's "CT480BX500SSD1 : 480.1 GB" - no maker name.
Googling the first part of that suggests it's Crucial )and 34.99 pounds
new via Amazon). So I'm getting the "Crucial storage executive". (It virus-scanned OK, but during install "Created with an evaluation version
of InstallBuilder", which doesn't fill one with confidence!)

I thought I'd check the warranty while there; I typed in what Crystal
said it was (2115E596FF67), and it (the website) said "SSD Serial Number
is invalid.". I clicked on "How to find my serial number?", an it told
me 1. on your product, 2. on the packaging, 3. via software - which I
chose, which led me to the "storage executive", which I have installed
and run - and it said 2115E596FF67! (Nevertheless, I copied-and-pasted
it from there into the website - still says it is invalid, of course.)
[So I've filled in a ticket, saying more or less what I have in this paragraph.]>

While the SSDs have some common entries, you can find vendor-specific
entries on a lot of them. And we don't know what the vendor-specific
entries mean. It's so bad in fact, that some drives, the Toolkit

Well, the "executive" has a button for the SMART data, and gives values
and titles for lots of them - though no guidance as to which are good or
bad (though overall says the drive is healthy).

does not support several of the drive models. They are that non-standard
and unsupported. That means the drive manufacturer (!) cannot display
the SMART, and quite possibly, the CrystalDiskInfo designer doesn't
know the answer either. The drive manufacturer likely does know,
but the info may be covered by an NDA preventing the sharing of the
info with mere customers.

HDTune 2.55 is perfectly safe as far as I know. If any info displayed

So I could try the left tab without shortening the SSD's life? (I had
assumed that did a read/write to all sectors.)

is not to your liking, the software is quite old and a lot of things
have happened since that free version was frozen. The supported (paid) version is much nicer looking and supports R/W characterization for
things like cheesy SMR (shingled HDD) drives. The read and write speeds
do not need to be identical on HDD, and that would happen on an SSD only
if the error corrector ARM cores can't keep up with the TLC read errors :-)

TLC or QLC drives can get "mushy" if left unpowered for periods of time.
For drives that are powered irregularly but at some frequency, the
SSD drives like that have maintenance routines they can run to freshen
the storage (at the cost of one cell wear for the affected areas).

I would only buy one and keep it that way intending to use it from new
(i. e. when this one dies), so would accept it doing a lengthy self-test
to get rid of "mushiness", if they can do that. But from what you're
saying, I don't need to do so for a while yet.

The "uncorrected" error information, helps the drive figure out how
mushy things are. For example, sectors in the free pool, would not
need to be "tended", as they don't contain user info. So if I had
a 4TB drive with 1TB of partitions, only 1TB of it would be a candidate
for "freshening". Windows 10 running TRIM once a week, helps inform the

Does it do that by default? I've never told it to; I would hope the shop
that sold me the laptop would have, though.

They're trying to make SATA disappear (the 870E boards only have four
SATA ports, versus the traditional six SATA ports). I think it is a reasonably estimate, that "some day", SATA will disappear just as
surely as IDE. But there is really no need of doing that. SATA is
a pretty economical standard, and they should always be able to support it.

Hmm. I suspect if (when) these laptops (7-32 and 10-64) finally die,
it'll be a major new system anyway, so not worrying about that.>

You should look at the prices of NAND, predictions of the future,
before buying a stocking stuffer for later. At some previous time,
I was warning people that at the time, that was a PERFECT TIME
to buy an SSD. But that time has passed and the price is a bit high
at the moment. If you find drives at lesser prices (like the NS100 drives
I used for scratch OS installs), those are not recommended for long
term storage, just because of how the latest ones have presented themselves

I would only be buying one against this one failing. For long-term
storage, I still feel happier with spinning drives, shingling
notwithstanding.

[]

The GPT issue with Windows 7, might have been with respect to booting
or something. I would not particularly expect a limitation on
only being able to use 2TB drives on Windows 7 x86. If you wanted
to use a 4TB GPT drive for data on Windows 7 x86 that should work.

Ah, I had a memory of there being a 2T limit _somewhwere_ - maybe it's partition size - for 7-32.

[]

And when cloning a drive, I think you can see the potential for trouble,
when taking a <2TB drive content and placing it on a >2TB device.

I had something similar on a previous generation: when restoring an
image from the drive that seized (I think it was a 1xx GB drive, and I
restored to a bigger one - might have been 2xx), I first restored making
a partition of the size I'd had before, then in Windows enlarged it. (I
know Macrium could have resized while restoring, but I felt happier
doing one thing at a time!) That may not have involved hardware/OS
limits, though.

The former drive may be MBR partitioned, the new drive may need to be
GPT, yet the cloning softwares may not be equipped for the "transition".
And partition manager softwares, may not have good coverage for this
case either. While Windows has an "MBR2GPT.exe" utility, that is
mainly intended for boot drive transitions and not for general
purpose partition handling. I've not tested that in a VM to see
what kind of a mess it makes of data drives. That's a research topic
for somebody.

The only PM that I've used in the last few years is the EaseUS one (and
the one in Windows, switching to the EaseUS one when the Windows one
couldn't do what I wanted - probably a C: shrink). I did use another one
years ago - I think it was just called Partition Manager. I can't
remember why I stopped using that - maybe it hit some limit, or maybe I
just didn't have it when I needed one, and the EaseUS one was the first
one I found, and as that did what I wanted, I didn't look further.>

Paul

John
--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

"To YOU I'm an atheist; to God, I'm the Loyal Opposition." - Woody Allen
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.windows7.general

Op 21-9-2025 om 20:52 schreef J. P. Gilliver:

I thought I'd check the warranty while there; I typed in what Crystal
said it was (2115E596FF67), and it (the website) said "SSD Serial Number
is invalid.". I clicked on "How to find my serial number?", an it told
me 1. on your product, 2. on the packaging, 3. via software - which I
chose, which led me to the "storage executive", which I have installed
and run - and it said 2115E596FF67! (Nevertheless, I copied-and-pasted
it from there into the website - still says it is invalid, of course.)
[So I've filled in a ticket, saying more or less what I have in this paragraph.]>

Just for fun, I checked your serial number and if it breaks you're out
of luck:

This product is valid but warranty has expired and this part is no
longer eligible for return

Crucial BX500 480GB 3D NAND SATA 2.5-inch SSD
Serial number
2115E596FF67
Part number
CT480BX500SSD1

Possibly you disabled a bit too much on the website.

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.windows7.general

On 2025/9/29 1:14:23, dillinger wrote:

Op 21-9-2025 om 20:52 schreef J. P. Gilliver:

I thought I'd check the warranty while there; I typed in what Crystal
said it was (2115E596FF67), and it (the website) said "SSD Serial Number
is invalid.". I clicked on "How to find my serial number?", an it told
me 1. on your product, 2. on the packaging, 3. via software - which I
chose, which led me to the "storage executive", which I have installed
and run - and it said 2115E596FF67! (Nevertheless, I copied-and-pasted
it from there into the website - still says it is invalid, of course.)
[So I've filled in a ticket, saying more or less what I have in this
paragraph.]>

Just for fun, I checked your serial number and if it breaks you're out
of luck:

This product is valid but warranty has expired and this part is no
longer eligible for return

Crucial BX500 480GB 3D NAND SATA 2.5-inch SSD
Serial number
2115E596FF67
Part number
CT480BX500SSD1

Possibly you disabled a bit too much on the website.

I filled out a ticket, and got a sort of standard (and thus irritating)
reply, asking for all the standard things - when I bought it, proof of purchase, etc.; I was in the process of replying saying that the problem
wasn't at my end but with the website, about to tell them _exactly_ what
the error message was that the website produced (maybe send them a
screenshot) - but the website then worked, as you have found. So,
although they didn't admit anything in their message, they'd obviously
fixed it. (I wasn't surprised to find my drive was out of warranty.)
--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()ALIS-Ch++(p)Ar++T+H+Sh0!:`)DNAf

An act like Morecambe and Wise happens once in a lifetime. Why did it
have to happen in mine?
- Bernie Winters quoted by Barry Cryer, RT 2013/11/30-12/6
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.windows7.general

On 9/21/2025 2:23 AM, J. P. Gilliver wrote:

On 2025/9/21 4:0:59, Paul wrote:

[]

Purely for your amusement, you could try a bad block scan of the drive.
This is pretty old software itself, having been released in 2008. The last >> column over will check whether the blocks have bad CRC. That version of
HDTune, does not do writes, so it should not make the SSD health any worse. >>
https://www.hdtune.com/files/hdtune_255.exe

<MAJOR SNIPING>

I thought that reading a "unit" on an SSD caused tits contents to be
rewrote in another unit and the original to go in a pool (over
partition?) that is used for wear leveling So whether that utility
writes or not, the SSD will have its pot stirred.

Is my memory wrong about this?
--
Jeff Barnett

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.windows7.general

On Mon, 9/29/2025 3:20 AM, Jeff Barnett wrote:

On 9/21/2025 2:23 AM, J. P. Gilliver wrote:

On 2025/9/21 4:0:59, Paul wrote:

[]

Purely for your amusement, you could try a bad block scan of the drive.
This is pretty old software itself, having been released in 2008. The last >>> column over will check whether the blocks have bad CRC. That version of
HDTune, does not do writes, so it should not make the SSD health any worse. >>>
-a-a-a https://www.hdtune.com/files/hdtune_255.exe

<MAJOR SNIPING>

I thought that reading a "unit" on an SSD caused tits contents to be rewrote in another unit and the original to go in a pool (over partition?) that is used for wear leveling So whether that utility writes or not, the SSD will have its pot stirred.

Is my memory wrong about this?

When a drive is TRIMmed, the OS sends a list of LBAs known to
not be used, and those can be put in the Free Pool and prepared
for usage. The preparation for usage is an Erase, which costs
one wear life. Any voltage changes in a NAND floating gate, can
cause small changes in crystal structure. Annealing a NAND cell
could fix these, extending life significantly, but we don't
know how to do that on a massive scale.

When you write a file, on a TLC or QLC device, the write method
can use an SLC cache (floating gate cell made to hold only one
bit, by changing the voltage threshold scheme), it costs you
one write for the SLC cache, and one write for the final placement
into TLC storage. The SLC cache is not a dedicated static thing,
it can be moved around by applying different voltage sensing
to any LBAs declared as being part of the cache area. It's not
like the critical-data-storage area, which might be an SLC area
at all times.

But reading is "free". It isn't DRAM and does not need refresh
in that sense. DRAM dribbles in milliseconds. The floating gate
in that case, is not protected by quantum mechanics.

On something like QLC, just about every sector read needs correction.
A 512 byte sector has 4KBits and the voltage thresholds are pretty tight.
If a read has one bit in error, this is easily corrected by Reed Solomon without needing the sector to be rewritten. The error corrector code
is quite strong and no reason to panic.

Say you read a sector and 20 bits out of 4096 are errored. Maybe the code
is still strong enough for this, correcting it. The syndrome is 50 bytes
on a 512 byte sector, to give some idea of the syndrome size.

But at some point, if you were tracking the statistics on the whole
page, as the firmware, you might decide to re-write the page by pulling
a new one from Free Pool, and copying the data over. These activities
are not normally described in any detail, so we are not privy to how
the devices are tuned.

But reading as an activity, does not need refresh in the same way
that DRAM needs constant refresh and lost bandwidth. In the DRAM,
there are refresh counters that go up through the address space,
and, in effect, have "timed" refresh activity. The DRAM memory
might be refreshed 64 times within the discharge period, so the
refresh is done in such a way, there is no possibility of a
discharge failure. In years past, people would fool with the
DRAM refresh settings, to "push their luck" :-)

Compare that to a floating gate NAND, where electrons leaving
the insulated gate is quantum mechanically disallowed. Maybe an
electron tunnels to escape, and the probability is quite low.
A NAND cell should be able to remain charged to a particular
voltage for ten years. In an SLC cell, the voltage threshold
is at 50%, and there are huge margins around the threshold.
The SLC NAND, as a beast, is hardly likely to be "mushy" in
everyday usage, and will be robust right up to the ten year
estimate. Whereas QLC, where there are 15 thresholds for the
16 voltage regions, that is tight-as-a-drum and the voltage
could shift in the cell, getting dangerously close to moving
into the next threshold region. We can correct a lot of these
shifts, with our powerful error corrector, but it's getting
harder to guarantee, by trickery, that we have robust
storage (with that correction scheme).

There was a Samsung, where it was noticed the read-performance
after three months, was dropping to 300MB/sec from 540MB/sec
and this was thought to be error corrector activity on
everything being read. And it is possible at that point,
that the re-write tuning was added or tuned, to stop users
from complaining. But we really don't know what was going
on there, in terms of details. This is all just guesses from
enthusiast web site news releases. Some SSD controller chips
have four cores now, one would be the command interpreter,
while three cores could be running error correction. That's
to help hide our precious purchase is mushy, do the correction
really fast...

In decades passed, there were reports that reading did wear
on cells. However, there is no anecdotal evidence of this
being anywhere near significant, in end-user usage patterns.
You can read the shit out of them, and if there is an
effect (like it were to be making the flash cells mushy),
users do not notice that a high-read application was
causing poorer eventual read performance, than SSDs
that are loafing along through their workday. If there is
a mechanism there, no one has joined the dots and claimed
to have seen this in practice.

I have never seen a comment from a person knowledgeable on the subject, regarding why there were early reports of an effect, and no such
claims are made today. The disconnect remains unexplained.
I get the impression though, that the mushy cases, they
can be mushy just while leaving the SSD sitting on the shelf
for three months, that particular device might have shown mush.
Whether mush can be cured by making the gate to channel
thicker, I don't know if that is a factor or not. Considering
the Z-axis and the 3D nature of NAND, there is some amount
of design pressure to make then thin-as-can-be. There can be
16 die inside a NAND chip, and then a couple hundred cells
in a column within each die. And yet the chips never seem
to grow in stature.

But on the other hand, advancement of NAND is kinda slow now.
They are working on five-bit cells (which means even tighter
thresholds), but those damn things have not poked their
little heads into super-cheap product. They're not shipping
yet. Sooner or later the write life is going to have to drop
below 600 or so, writes per cell.

Micron makes an industrial flash with 6x the write life, but
like everything in high tech, the price has to be 6x higher.
Maybe you might find those in certain Enterprise products.
The 600 number is not some "manifest constant", it is just
a common number for consumer devices. There have been devices
with considerably better numbers. And if they still made actual
true SLC devices, those could manage 100,000 writes per cell,
and would outlast your lifetime. Making 600 ones, ensures,
like bog roll, there is a kind of rent-seeking behavior.
But 600 still lasts a long time when all you do is web surf
(and you set your web cache to be in RAM :-) ).

Paul
--- Synchronet 3.21a-Linux NewsLink 1.2