Are HDD manufacturers "sorting" drives by quality before chosingwhich is sold as which product?

[Daniel Smedegaard Buus]

Hello

I have a mighty large ZFS array (19*2TB) which has two drives that are
currently having a few too many sector reallocations for my taste. So,
I went looking for replacements. These days, drives are about twice as
expensive as last I bought any. But one drive is really cheap
(apparently 50% off right now in a Danish store) - this one:
http://www.newegg.com/Product/Product.aspx?Item=22-145-326

Thing is, this product has received extremely bad reviews, both on
Newegg and Amazon. Tons of DOA, tons of drives dying within minutes,
hours, and days. Which is puzzling, because apparently, inside it is
this drive: http://www.newegg.com/Product/Product.aspx?Item=N82E16822145473
- which is not receiving the same kind of bad reviews.

So, I'm thinking either these drives are refurbished (to which I
cannot find any hard or circumstantial evidence), or something else is
fishy. Then, I'm remembering that I myself once bought WD external
drives, because they were cheaper than their internal counterparts,
although the two products' drives were identical. I just thought,
"Lucky me, I get to save money AND get a free external case!". Those
drives have proven to be less than excellent. In fact, the drives I'm
about to replace, are two of these. With only about 240 days of
uptime.

What I'm thinking now is that some kind of selection might be going on
in the HDD plants based on some low-level testing. Like when Intel or
AMD select parts from the same wafer to be sold as differently clocked
versions of a processor, simply because the quality of a wafer isn't
100% consistent. Perhaps there's a quality assessment going on in
Hitachi's plants that will provide estimates on a particular drive's
expected lifespan?

Such an estimate would make sense to be used as a basis for selecting
internal vs. external drives, simply because external drives
supposedly would be used less frequently, and "only" for backup. So,
their lifespan would be expected to pretty much match up with better
rated drives that are sold as internal drives and used more
extensively.

This, of course, is all speculation on my part, which is why I'm
posting here. Does any one of you have enough insight to either
confirm or deny my suspicions?

Thanks in advance,
Daniel

 

[Rod Speed]

No they are not.

I have a mighty large ZFS array (19*2TB) which has two drives that
are currently having a few too many sector reallocations for my taste.

Likely you arent cooling them adequately.

So, I went looking for replacements. These days, drives
are about twice as expensive as last I bought any.

That's because of the floods in Thailand affecting the HDD factorys there.

But one drive is really cheap (apparently 50%
off right now in a Danish store) - this one:
http://www.newegg.com/Product/Product.aspx?Item=22-145-326

The reason its so cheap is because its only a USB2 drive.

There are hordes of USB3 drives around now.

Thing is, this product has received extremely bad reviews, both on
Newegg and Amazon. Tons of DOA, tons of drives dying within minutes,
hours, and days. Which is puzzling, because apparently, inside it is this
drive:
http://www.newegg.com/Product/Product.aspx?Item=N82E16822145473
- which is not receiving the same kind of bad reviews.

Likely it isnt been cooled adequately in that enclosure.

So, I'm thinking either these drives are refurbished

Very unlikely.

(to which I cannot find any hard or circumstantial evidence),
or something else is fishy.

Likely the drive just isnt being cooled adequately.

Then, I'm remembering that I myself once bought WD external
drives, because they were cheaper than their internal counterparts,
although the two products' drives were identical. I just thought,
"Lucky me, I get to save money AND get a free external case!". Those
drives have proven to be less than excellent. In fact, the drives I'm
about to replace, are two of these. With only about 240 days of uptime.

Likely because they arent being cooled adequately.

Post the SMART data on those drives.

What I'm thinking now is that some kind of selection might be
going on in the HDD plants based on some low-level testing.

Unlikely given that they would have to be replaced under warranty.

Like when Intel or AMD select parts from the same wafer to
be sold as differently clocked versions of a processor, simply
because the quality of a wafer isn't 100% consistent.

Hard drives are entirely different technology.

Perhaps there's a quality assessment going on
in Hitachi's plants that will provide estimates
on a particular drive's expected lifespan?

MUCH more likely they just arent being cooled
adequately in that particular enclosure.

Such an estimate would make sense to be used as a basis for
selecting internal vs. external drives, simply because external drives
supposedly would be used less frequently, and "only" for backup.

That's not right. They are used extensively with PVRs.

So, their lifespan would be expected to pretty much match up with better
rated drives that are sold as internal drives and used more extensively.

That line cant fly either.

This, of course, is all speculation on my part, which is
why I'm posting here. Does any one of you have enough
insight to either confirm or deny my suspicions?

Your theory cant fly. And you can prove that it's the
temperature that's the problem with the SMART data.

 

[Daniel Smedegaard Buus]

No they are not.

Would you care to elaborate on your conclusion? Do you work in the
industry, know someone, have links to facts on the matter, etc.?

Likely you arent cooling them adequately.

Off-topic, but: Likely based on what? I monitor everything in my
machine, smartctl informs me of any irregularities. It's summer right
now, and the two drives in question are 36 and 40 degrees C,
respectively. I have a very nice airflow in my rig.

That's because of the floods in Thailand affecting the HDD factorys there..

I know.

The reason its so cheap is because its only a USB2 drive.

Because adding a case, a PSU, and a USB 2.0 port to an otherwise bare
$175 harddrive makes it a better sell at half the original price?

There are hordes of USB3 drives around now.

Good to know.

Likely it isnt been cooled adequately in that enclosure.

Likely based on what knowledge?

Very unlikely.

Very unlikely based on what knowledge?

Likely the drive just isnt being cooled adequately.

Likely based on what knowledge?

Likely because they arent being cooled adequately.

Likely based on what knowledge?

Post the SMART data on those drives.

No. Off-topic.

Unlikely given that they would have to be replaced under warranty.

Only if they break within the warranty period.

Hard drives are entirely different technology.

As are kitchen knives, diamonds, RAM modules, shoes, meat, etc., and
yet comparable sorting strategies are used with those products.

MUCH more likely they just arent being cooled
adequately in that particular enclosure.

Based on what knowledge?

That's not right. They are used extensively with PVRs.

Sure, some will be used with PVRs. Tons of other use cases too, but
I'd still believe that external drives are used less intensively than
internal ones.

That line cant fly either.

Well, as I've made clear, it's all speculation on my part, but if we
can agree that a drive has a statistical X hours of use in it, then we
should be able to agree that one that's used 24/7 is statistically
going to die before one that's used a couple of hours a week.

Your theory cant fly.

Based on what knowledge?

And you can prove that it's the
temperature that's the problem with the SMART data.

Well, I could "prove" that if that were the case, but it's not. So I
can't.

So, okay, if you're actually interested in helping with the question
based on actual knowledge or experience, I welcome you to, and I will
be grateful for any insight you may bring. But if you're just looking
for an argument, which your repetitive one-liners seem to suggest,
please just leave it be.

Thank you.

 

[Yousuf Khan]

Hello

I have a mighty large ZFS array (19*2TB) which has two drives that are
currently having a few too many sector reallocations for my taste. So,
I went looking for replacements. These days, drives are about twice as
expensive as last I bought any. But one drive is really cheap
(apparently 50% off right now in a Danish store) - this one:
http://www.newegg.com/Product/Product.aspx?Item=22-145-326

Thing is, this product has received extremely bad reviews, both on
Newegg and Amazon. Tons of DOA, tons of drives dying within minutes,
hours, and days. Which is puzzling, because apparently, inside it is
this drive: http://www.newegg.com/Product/Product.aspx?Item=N82E16822145473
- which is not receiving the same kind of bad reviews.

Yeah, I've stopped buying external drives as a rule now. I still have a
some enclosures left over, some of which I still use as a external
drives for PVR's, or as lesser used PC drives.

So, I'm thinking either these drives are refurbished (to which I
cannot find any hard or circumstantial evidence), or something else is
fishy. Then, I'm remembering that I myself once bought WD external
drives, because they were cheaper than their internal counterparts,
although the two products' drives were identical. I just thought,
"Lucky me, I get to save money AND get a free external case!". Those
drives have proven to be less than excellent. In fact, the drives I'm
about to replace, are two of these. With only about 240 days of
uptime.

I think these external enclosures are murder for these drives,
especially when most of them don't have anything but passive cooling.
But I've seen massive failures even in actively cooled enclosures.

Yousuf Khan

 

[Rod Speed]

Would you care to elaborate on your conclusion?

I did. What you are seeing is most likely just lack of
cooling, not deliberate use of drives that wont last long.

Do you work in the industry,
Yes.

know someone, have links to facts on the matter, etc.?

I told you how to see if the failures you are seeing
is due to the drive not getting adequate cooling.

Off-topic,

Nope, completely on topic.

but: Likely based on what?

Based on what has happened to many drives that have
ended up with a lot more reallocated sectors than normal.

Corse since you werent specific about the
actual numbers with the drives you believe
have too many of those, its possible that
the numbers arent exceptional.

I monitor everything in my machine, smartctl informs
me of any irregularities. It's summer right now, and
the two drives in question are 36 and 40 degrees C,
respectively. I have a very nice airflow in my rig.

Post the actual SMART data for those drives you believe
need to be replaced.

I know.

Because adding a case, a PSU, and a USB 2.0 port to an otherwise bare
$175 harddrive makes it a better sell at half the original price?

It is when they want to get rid of the USB2 drives
because no one will buy them if you try to sell
them for the same price as the USB3 drives.

Good to know.

Likely based on what knowledge?

What happens with many external drives. Its completely trivial
to see what temperature the drive gets to in use with SMART.

Very unlikely based on what knowledge?

It isnt even legal to sell those drives with refurbished drives in them.

And there just arent enough refurbished drives to do that anyway.

Likely based on what knowledge?

How drives die in external enclosures.

Its normally either lack of adequate cooling or getting dropped etc.

That's why for a long time the warrantys on external drives
were only 1 year with bare drives having 3 or 5 year warrantys.

Likely based on what knowledge?

How drives die in external enclosures.

Its normally either lack of adequate cooling or getting dropped etc.

That's why for a long time the warrantys on external drives
were only 1 year with bare drives having 3 or 5 year warrantys.

No. Off-topic.

It isnt off topic. That data will show what temperatures the drives
have seen and whether the number of reallocated sectors is unusual.

Only if they break within the warranty period.

And you claimed that they mostly do with the drives you mentioned first.

As are kitchen knives, diamonds, RAM modules, shoes, meat, etc.,
and yet comparable sorting strategies are used with those products.

Nope, nothing like it predicted life wise.

Based on what knowledge?

How drives die in external enclosures.

Its normally either lack of adequate cooling or getting dropped etc.

That's why for a long time the warrantys on external drives
were only 1 year with bare drives having 3 or 5 year warrantys.

Sure, some will be used with PVRs. Tons of other use cases too,

So that last bit of yours is just plain wrong.

but I'd still believe that external drives are
used less intensively than internal ones.

Sure, but that's not what you actually said with the last bit.

Well, as I've made clear, it's all speculation on my part, but if
we can agree that a drive has a statistical X hours of use in it,
then we should be able to agree that one that's used 24/7

That doesn't happen with many internal drives. Only with servers
that don't use the particular bare drive you listed much at all.

is statistically going to die before one
that's used a couple of hours a week.

That is not in fact the case with internal drives that are adequately
cooled.

Based on what knowledge?

How drives die in external enclosures.

Its normally either lack of adequate cooling or getting dropped etc.

That's why for a long time the warrantys on external drives
were only 1 year with bare drives having 3 or 5 year warrantys.

Well, I could "prove" that if that were the case, but it's not. So I
can't.

So, okay, if you're actually interested in helping with the question
based on actual knowledge or experience, I welcome you to,

You in fact refused to provide the SMART data which will show
whether your 'taste' about reallocated sectors is out of whack
of whether there is a real problem with those two drives.

and I will be grateful for any insight you may bring.
But if you're just looking for an argument, which
your repetitive one-liners seem to suggest,

Corse you never do anything like that yourself, eh ?

please just leave it be.

Request denied.

 

[Ed Light]

Just filter Roddy out. He's our troll.
--
Ed Light

Better World News TV Channel:
http://realnews.com

Iraq Veterans Against the War and Related:
http://ivaw.org
http://couragetoresist.org
http://antiwar.com

Send spam to the FTC at
(e-mail address removed)
Thanks, robots.

 

[Ed Light]

I think these external enclosures are murder for these drives,
especially when most of them don't have anything but passive cooling.
But I've seen massive failures even in actively cooled enclosures.

Yousuf Khan

You can use HD Sentinel, unregistered, to monitor the HD temps. If one
of those cases is unventilated, you can drill some holes in it.

In my totally unventilated case, my WD Green EAVS only gets a bit warmer
than I'd like under sustained copying. I'm thinking about the drill.

What I hate about external drives is encountering one of those
motherboards that go haywire when you reboot with a USB HD attached.


--
Ed Light

Better World News TV Channel:
http://realnews.com

Iraq Veterans Against the War and Related:
http://ivaw.org
http://couragetoresist.org
http://antiwar.com

Send spam to the FTC at
(e-mail address removed)
Thanks, robots.

 

[Yousuf Khan]

You can use HD Sentinel, unregistered, to monitor the HD temps. If one
of those cases is unventilated, you can drill some holes in it.

Yup use that already, registered.

In my totally unventilated case, my WD Green EAVS only gets a bit warmer
than I'd like under sustained copying. I'm thinking about the drill.

I got one drive, from Lacie, which hits 60C during copying. It then just
shuts itself down after awhile. I can't even reliably mass copy stuff
*off* of it, as that also overheats it. And since it's a USB 2.0
connection, it comes off of it slowly, thus prolonging the overheating
period. The data that's already in there is in good shape though, so far.

What I hate about external drives is encountering one of those
motherboards that go haywire when you reboot with a USB HD attached.

I got one of those motherboards. At its mildest, the BIOS just changes
default boot drive when one of these external drives is attached. At its
worst, the computer refuses to boot up if the external drive is attached.

Yousuf Khan

 

[Yousuf Khan]

Well, I could "prove" that if that were the case, but it's not. So I
can't.

So, okay, if you're actually interested in helping with the question
based on actual knowledge or experience, I welcome you to, and I will
be grateful for any insight you may bring. But if you're just looking
for an argument, which your repetitive one-liners seem to suggest,
please just leave it be.

Thank you.

I've tried to get intelligent responses out of Rod before too. It
occasionally works, and he gives out good information, but then he goes
off into Rod world again within a couple of sentences. It's not worth
sorting though this haystack of nuttiness for the occasional good needle
of information. Just filter him out now.

Yousuf Khan

 

[larrymoencurly]

What I'm thinking now is that some kind of selection might be going on
in the HDD plants based on some low-level testing. Like when Intel or
AMD select parts from the same wafer to be sold as differently clocked
versions of a processor, simply because the quality of a wafer isn't
100% consistent. Perhaps there's a quality assessment going on in
Hitachi's plants that will provide estimates on a particular drive's
expected lifespan?

Such an estimate would make sense to be used as a basis for selecting
internal vs. external drives, simply because external drives
supposedly would be used less frequently, and "only" for backup. So,
their lifespan would be expected to pretty much match up with better
rated drives that are sold as internal drives and used more
extensively.

I once saw a photo of a 5.25" internal hard disk (20MB?) with a sticker on it that said, "IBM REJECT". I think it was shown in a 1980s PC magazine, and the HD had been sold as brand new by a mail order retailer. Is that thekind of HD you have?

I've read that Fantom external drives sold as brand new sometimes contained"recertified" internal drives, and in 2005 this Amazon customer said his Fantom included an internal labelled "Magnetic Data Technologies", a repair company.

http://www.amazon.com/review/R5P35TWY1WU1E/ref=cm_cr_pr_viewpnt#R5P35TWY1WU1E

Here's a thread about Magnetic Data Technologies:

http://forums.storagereview.com/index.php/topic/13679-magnetic-data-technology/

OTOH I doubt manufacturers of internal HDs favor their lower quality products for their external HDs because a couple of years ago, some WD 2TB Elements USB 2.0 drives were shipped with high performance Caviar Black 7200RPM drives inside, while others had only 5400RPM Caviar Green drives in them. These external drives were all the same model, but the particular internal drive could be identified through some markings on the cardboard box (threadabout this at FatWallet.com, about a Black Friday sale at Target or Staples). If WD wanted to dump its worst products into its externals, why would it bother putting costlier 7200RPM drives in them when USB 2.0 can't benefit from the higher RPMs or bigger cache?

 

[Oscar]

I've tried to get intelligent responses out of Rod before too. It
occasionally works, and he gives out good information, but then he goes
off into Rod world again within a couple of sentences. It's not worth
sorting though this haystack of nuttiness for the occasional good needle
of information.

Says the fool that keeps getting done like
a ****ing dinner, time after time after time.

Just filter him out now.

Fat lot of good that will ever do you.

And NONE of you clowns even commented on his original anyway.

 

[Daniel Smedegaard Buus]

I've tried to get intelligent responses out of Rod before too. It
occasionally works, and he gives out good information, but then he goes
off into Rod world again within a couple of sentences. It's not worth
sorting though this haystack of nuttiness for the occasional good needle
of information. Just filter him out now.

        Yousuf Khan

Thanks, I read a couple of the lines he responded with the second time
around and could see it wasn't going in any good direction, so just
skipped past it.

About the enclosure part of it, I definitely agree that a tiny box
isn't a place for anyone to live in, let alone a harddrive, let alone
a 7200 RPM one of the kind

However, in my particular case, I bought 13 of those and stripped them
of their boxes immediately. They've been living happy lives ever since
in temps ranging betwen over 20 C and under 50 C. And yet, one of them
actually died to the point where the WD LifeGuard util said to replace
it under warranty after less than six months. These other two, with
200-some days on them both have in the hundreds of reallocated
sectors, with new ones showing up after either a long or short
selftest. I'm pretty used to seeing LBAs for sectors that were read
erroneously and caused a SMART test to fail, but usually it's just a
question of doing one, maybe two writes to bad LBAs, then a subsequent
long selftest will succeed. This time around, I don't think they're
coming back. 24 hours of wash-rinse-repeat and they're still failing
selftests with new pending sector allocations.

Anyway, back on topic — those drives I'm linking to most definitely
would be considered likely to get pretty warm, even if there are
"ventilation holes" in the cases. They *are* 7200 RPM, after all. But
still, so many reports of DOA and dead within *days* or weeks, just
doesn't seem to add up. My first frail attempt at doing hobby raid
some ten years ago was a mix-and-match of old cheap odd-sized 7200 RPM
ATA drives in a cramped miditower, running Linux and md RAID5. I never
monitored the fans and I don't know how long they'd been dead when
pretty much every disk was dying in a flaming hell. A long time, I'm
guessing, as not just one, but two fans had eventually set out, and I
literally scolded my fingers on one of the drives. LOL

Still, I was able to salvage about half of my data on that rig, and
while the drives were dying, they never set out per se — it was "just"
bad sector hell. I actually still have some of those drives. I like to
think of them as a place to put bad films that I know I should just
delete but that I can't bring myself to do so to

Just makes me think that hours or days in an external case with
ventilation holes, while not ideal for a 7200 RPM SATA drive,
shouldn't be able kill it off completely. It doesn't add up in my
head

 

[Daniel Smedegaard Buus]

On Friday, June 1, 2012 11:24:35 PM UTC-7, Daniel Smedegaard Buus wrote:

I once saw a photo of a 5.25" internal hard disk (20MB?) with a sticker on it that said, "IBM REJECT".  I think it was shown in a 1980s PC magazine, and the HD had been sold as brand new by a mail order retailer.  Is that the kind of HD you have?  

LOL No, unfortunately not, that would've been pretty cool

I've read that Fantom external drives sold as brand new sometimes contained "recertified" internal drives, and in 2005 this Amazon customer said hisFantom included an internal labelled "Magnetic Data Technologies", a repair company.

 http://www.amazon.com/review/R5P35TWY1WU1E/ref=cm_cr_pr_viewpnt#R5P35...

Here's a thread about Magnetic Data Technologies:

http://forums.storagereview.com/index.php/topic/13679-magnetic-data-t...

Interesting. Here in Denmark, we've also had our share of stories of
people getting supposedly new computers that are full of some other
person's personal data. In a large store chain, Bilka, a few years
back they were actually selling hard drives that'd been refurbished —
I actually think "recertified" might've been the phrasing chosen by
the supplier.

It's exactly stuff like this that makes me post questions like this.

OTOH I doubt manufacturers of internal HDs favor their lower quality products for their external HDs because a couple of years ago, some WD 2TB Elements USB 2.0 drives were shipped with high performance Caviar Black 7200RPMdrives inside, while others had only 5400RPM Caviar Green drives in them.  These external drives were all the same model, but the particular internal drive could be identified through some markings on the cardboard box (thread about this at FatWallet.com, about a Black Friday sale at Target or Staples).  If WD wanted to dump its worst products into its externals, why would it bother putting costlier 7200RPM drives in them when USB 2.0 can't benefit from the higher RPMs or bigger cache?

Well, actually, that's *exactly* my point. Supposing that WD in this
case didn't specifically make an "inferior" product to be sold as
external drives. Of all the external drives I've owned, what's inside
can also be bought as a bare drive. Rather, what if they analyze their
"yield", prune out the worst of the bad apples and recycle them for
parts, then prune out the better bad apples from all manufacturing
lines and chose those for external drives. That would explain exactly
why they'd put a Caviar Black inside an external enclosure — because
it was of lower quality than the rest of that batch of Caviar Blacks.
I mean, assuming they wouldn't make mistakes like that (I assume they
make *no* mistakes at all), then why wouldn't they just put Caviar
Greens in all of them? If it's not a mistake, then there's a reason.
It *could* be what I'm suggesting. It could also be that they had
obligated themselves to deliver a certain order and production
couldn't keep up so they had to bite the bullet. It could also be that
the difference between one type of drive and another is really
controlled by the firmware, so they make the same kind of money either
way.

What I'm saying is that any kind of tech production that I can think
of produces a more or less inconsistent quality. At Intel and AMD, the
lesser quality wafer parts are used for lower-clocked, lower-powered
and/or core-disabled parts. At Nvidia and ATI the best wafer parts go
in the fastest graphics cards. At Crucial the best chips are rated for
the highest-clocked memory modules.

Assuming that production of a specific type of harddrive is a custom
process, what happens to the lower-rated output of such a process? I
refuse to believe that such a high-tec business isn't able to rate
each drive based on multiple parameters. We already know they test
each and every drive. That's where the plist is populated.

Again, I'm speculating, but if it makes sense financially, it makes
sense in business. If the auto industry can decide whether to withcall
a car from the market based on the price of that versus the estimated
price of lawsuits when people start dying from some defect, then
surely this isn't too crazy to a thought

 

[4783l]

LOL No, unfortunately not, that would've been pretty cool

Interesting. Here in Denmark, we've also had our share
of stories of people getting supposedly new computers
that are full of some other person's personal data.

Those clearly were not refurbished drives, just complete
systems returned by the purchaser and sold to someone else.

In a large store chain, Bilka, a few years back they were actually
selling hard drives that'd been refurbished — I actually think
"recertified" might've been the phrasing chosen by the supplier.

Different thing entirely to selling them as brand new.

It's exactly stuff like this that makes me post questions like this.

It cant fly, there arent enough refurbished drives for the manufacturer
to be selling all external drives with refurbished drives in them.

Well, actually, that's *exactly* my point.

No it is not.

Supposing that WD in this
case didn't specifically make an "inferior" product to be sold as
external drives. Of all the external drives I've owned, what's inside
can also be bought as a bare drive. Rather, what if they analyze their
"yield", prune out the worst of the bad apples and recycle them for
parts,

Costs more than making a new drive from scratch in asia.

then prune out the better bad apples from all
manufacturing lines and chose those for external drives.

There clearly arent enough of those for the external drives they
sell when WD had to include the Caviar Blacks in some of theirs.

That would explain exactly why they'd put
a Caviar Black inside an external enclosure

The other explanation is MUCH more plausible, they just didn’t
have enough of the Caviar Greens to satisfy the demand.

— because it was of lower quality than the rest of that batch
of Caviar Blacks. I mean, assuming they wouldn't make mistakes
like that (I assume they make *no* mistakes at all), then why
wouldn't they just put Caviar Greens in all of them?

Because they don’t have enough of them.

If it's not a mistake, then there's a reason.

Because they don’t have enough of them.

It *could* be what I'm suggesting.

Nope. There just arent enough refurbished
drives to put them in the external cases.

It could also be that they had obligated themselves
to deliver a certain order and production
couldn't keep up so they had to bite the bullet.

And that’s a MUCH more plausible explanation than your
conspiracy theory, particularly if they found they werent
selling a substantial percentage of the Caviar Blacks for
whatever reason.

It could also be that the difference between one type
of drive and another is really controlled by the firmware,
so they make the same kind of money either way.

We know it is not. The rotational speed is different.

What I'm saying is that any kind of tech production that
I can think of produces a more or less inconsistent quality.

Yes, but what matters is whether they get enough of the
duds to supply the entire external drive market. They
clearly don’t even get enough of the Greens to do that.

At Intel and AMD, the lesser quality wafer parts are used for
lower-clocked, lower-powered and/or core-disabled parts.

Yes, but hard drives don’t work like that except in the sense
that drives with one surface unviable can be sold as a drive
of the lower capacity in that model range.

At Nvidia and ATI the best wafer parts go
in the fastest graphics cards. At Crucial the best chips
are rated for the highest-clocked memory modules.

And with hard drives, the best drives are sold at the
maximum capacity for the model range and the ones with
less than perfect surfaces are sold as the smaller drives.

Assuming that production of a specific type of harddrive is a custom
process,

It isnt.

what happens to the lower-rated output of such a process?

They get sold as the smaller drives in the model range.

I refuse to believe that such a high-tec business isn't
able to rate each drive based on multiple parameters.

Corse they can.

We already know they test each and every
drive. That's where the plist is populated.

And the drives that have less than perfect
surfaces are sold as the drives of lower size.

Again, I'm speculating, but if it makes sense
financially, it makes sense in business.

Yes it does with drive size.

If the auto industry can decide whether to withcall a car
from the market based on the price of that versus the
estimated price of lawsuits when people start dying from
some defect, then surely this isn't too crazy to a thought

It is when they can just sell the less than perfect drives as smaller size
drives.

 

[Daniel Smedegaard Buus]

Bla bla bla

Not jumping into this Rod, sorry. Find someone else to troll. Not
gonna read it.

 

[Joseph Terner]

I have a mighty large ZFS array (19*2TB) which has two drives that are
currently having a few too many sector reallocations for my taste. So, I
went looking for replacements. These days, drives are about twice as
expensive as last I bought any. But one drive is really cheap
(apparently 50% off right now in a Danish store) - this one:
http://www.newegg.com/Product/Product.aspx?Item=22-145-326

Thing is, this product has received extremely bad reviews, both on
Newegg and Amazon. Tons of DOA, tons of drives dying within minutes,
hours, and days.

Almost all early failures happen due to transport damage. I bet the
retail package of the drive is inadequate for a hard disk and it's
handled like most consumer equipment without special care.

Joseph

 

[Daniel Smedegaard Buus]

Almost all early failures happen due to transport damage. I bet the
retail package of the drive is inadequate for a hard disk and it's
handled like most consumer equipment without special care.

Joseph

Could be. I just tried looking for some packagin dimensions vs. item
dimensions. Pretty hard to tell anything from it, though, but I did
find a PDF on archive.org,
http://web.archive.org/web/20110215...8825773F007FEE8C/$file/xl_desk_datasheet..pdf

Funny thing is - I had to go to archive.org because the product page
no longer exists on Hitachi's site. It's still here, though:
http://web.archive.org/web/20110112...t.com/external-drives/desktop/hitachi-xl-desk

And, it also just disappeared (today) from the local dealer:
http://www.computersalg.dk/pages/article.aspx?aID=17

Guess someone pulled the plug on that product - seems to make sense
whatever the reason is for it failing so

 

[4783l]

Not jumping into this Rod, sorry. Find someone else to troll.

You're the one trolling with your silly line about off spec drives
used in external enclosures.

Not gonna read it.

Its obvious you have done that, liar.

 

[4783l]

Could be. I just tried looking for some packagin dimensions vs. item
dimensions. Pretty hard to tell anything from it, though, but I did
find a PDF on archive.org,
http://web.archive.org/web/20110215...58825773F007FEE8C/$file/xl_desk_datasheet.pdf

Funny thing is - I had to go to archive.org because
the product page no longer exists on Hitachi's site.

Because Hitachi has flogged their hard drive operation.

It's still here, though:
http://web.archive.org/web/20110112...t.com/external-drives/desktop/hitachi-xl-desk

And, it also just disappeared (today) from the local dealer:
http://www.computersalg.dk/pages/article.aspx?aID=17

Guess someone pulled the plug on that product

Because that's the reason Hitachi's hard drive operation
was bought, so they can eliminate the competition.

 

[larrymoencurly]

Well, actually, that's *exactly* my point. Supposing that WD in this
case didn't specifically make an "inferior" product to be sold as
external drives. Of all the external drives I've owned, what's inside
can also be bought as a bare drive. Rather, what if they analyze their
"yield", prune out the worst of the bad apples and recycle them for
parts, then prune out the better bad apples from all manufacturing
lines and chose those for external drives. That would explain exactly
why they'd put a Caviar Black inside an external enclosure — because
it was of lower quality than the rest of that batch of Caviar Blacks.
I mean, assuming they wouldn't make mistakes like that (I assume they
make *no* mistakes at all), then why wouldn't they just put Caviar
Greens in all of them? If it's not a mistake, then there's a reason.
It *could* be what I'm suggesting. It could also be that they had
obligated themselves to deliver a certain order and production
couldn't keep up so they had to bite the bullet. It could also be that
the difference between one type of drive and another is really
controlled by the firmware, so they make the same kind of money either
way.

More likely, WD had a shortage of Green drives for their USB enclosures because those 2TB Black drives cost more to make, having 4 platters instead ofthe 3. I believe if WD wanted to get rid of lower quality HDs, instead ofselling them as externals they'd use them as warranty replacements, especially for warranties that were about to expire. Also when I've run MHDD, a program that reports slow sectors, the WD drives I tested showed no sectorsslower than 50ms, while the Hitachi, Samsung, and Seagate drives had some sectors that took 50-150ms and, 1-10 that took 150-500ms. Hitachi 1TB 7200RPM drives had none of the latter. 150ms is 13-18 rotations. I don't knowif the results were due to inferior heads or signal processing, more defects on the platters, or lower standards for screening out bad sectors. MHDDis available at HDDguru.com, which has forums full of HD experts who may be able to tell you if external drives are built from inferior internals.

What I'm saying is that any kind of tech production that I can think
of produces a more or less inconsistent quality. At Intel and AMD, the
lesser quality wafer parts are used for lower-clocked, lower-powered
and/or core-disabled parts. At Nvidia and ATI the best wafer parts go
in the fastest graphics cards. At Crucial the best chips are rated for
the highest-clocked memory modules.

But at least those chip companies label their products according to speed, voltage, and maximum temperature. OTOH memory modules sold on the retail market seem to vary widely in quality and are usually made from sub-prime oroverclocked chips. One example of this is shown in the following APHnetworks review of a G.Skill PC17000 Ripjaws module. Once the heatsinks were removed, Hynix PC10666 chips are revealed:

http://aphnetworks.com/reviews/g_skill_ripjaws_f3_12800cl7d_8gbrh_2x4gb/2

G.Skill is hardly alone in this practice. Module companies vary greatly inquality standards. Some test using US $1M+ machines, while most use just PC motherboards running at room temperature. At least one company even thinks it's OK for its modules to show 2 bad bits during such testing. For these and other reasons, it's best to buy modules only if they have no heatsinks (useless anyway) and contain chips clearly marked with the logo or partnumber of a real chip manufacturer. About the only way to be guaranteed of such modules if you don't visit the store is to buy Samsung or no-heatsink Crucial modules. Some people have reported that Samsung PC12800 modules can run faster than some heatsinked PC17000 modules. BTW, 1.35V DDR3 is rated to work in 1.5V motherboards.