Epox 8KHA+ - Maximum hard drive size?

[Dalamar]

Does anyone know, what is the biggest hard drive that the 8KHA+ can use?
Specifically, I mean the largest drive that the BIOS can handle. I don't want
to use an overlay driver.

Epox says that the largest they have tested is 80 gigs, but that doesn't tell
me if it can handle more than that.

Is anyone using this board with something bigger?

Thanks for any info anyone can provide.

 

[Lowtarget]

I'm using 8KHA+ my self. I'm have two hdd's installed in the system. A
120gig Western Digital and 80gig Western Digital. I don't have to use a
overlay driver at all. That mobo can handle a 120gig hdd very easy without
problems. I never any hdd's bigger then 120gig in the system. Then again I
don't own any hdd's bigger then 120gig's.

 

[Dalamar]

I'm using 8KHA+ my self. I'm have two hdd's installed in the system. A
120gig Western Digital and 80gig Western Digital. I don't have to use a
overlay driver at all. That mobo can handle a 120gig hdd very easy without
problems. I never any hdd's bigger then 120gig in the system. Then again I
don't own any hdd's bigger then 120gig's.

Thanks. I was looking at getting something of that size, at least, so it's
good to hear from someone who has one up and running.

 

[Robert Pendell]

The motherboard should be able to support 127GB before the address
translation problem kicks in. If your board support 48-bit address
translation, then the limit becomes alot higher but until then... If I am
wrong then someone correct me on this.

 

[Paul Taylor]

The motherboard should be able to support 127GB before the address
translation problem kicks in. If your board support 48-bit address
translation, then the limit becomes alot higher but until then...

ATA33/66/100 is limited to <= 137MB hard disks (in practice, 120MB.)
ATA133 and SATA can support > 137MB hard disks.

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Paul Taylor Veni, vidi, tici -
(e-mail address removed) I came, I saw, I ticked.

 

[Eric Gisin]

Nonsense. They are unrelated.

 

[Dalamar]

The motherboard should be able to support 127GB before the address
translation problem kicks in. If your board support 48-bit address
translation, then the limit becomes alot higher but until then... If I am
wrong then someone correct me on this.

Thank you for the help.

 

[Rod Speed]

Robert Pendell wrote:

ATA33/66/100 is limited to <= 137MB hard disks (in practice, 120MB.)

Wrong. There are hordes of those systems that dont have that limit.

ATA133 and SATA can support > 137MB hard disks.

And so can most of the others too.

 

[Marc de Vries]

Wrong. There are hordes of those systems that dont have that limit.


And so can most of the others too.

That just depends on the age of your mobo.

He is right that initially ATA100 boards were limited to 137MB. One of
the improvements of ATA133 was the support for larger disks.

But newer ATA100 devices also support larger disks now.
(Lots of companies didn't want to use ATA133 as defined by Maxtrox?
But they did see the need for larger disks)

So the correct answer is that you are sure that larger disks are
supported on ATA133, but you cannot be sure that larger disks are
supported on ATA100.

Since the 8KHA+ is a fairly old board, there is a good chance that it
is not supported.

I'm suprised at the terrible answer from Epox. They never tested with
larger then 80GB? They don't know the specifications of their own
boards?
The person at Epox that gave that answer should be ashamed of himself.

Marc

 

[Dalamar]

That just depends on the age of your mobo.

He is right that initially ATA100 boards were limited to 137MB. One of
the improvements of ATA133 was the support for larger disks.

But newer ATA100 devices also support larger disks now.
(Lots of companies didn't want to use ATA133 as defined by Maxtrox?
But they did see the need for larger disks)

So the correct answer is that you are sure that larger disks are
supported on ATA133, but you cannot be sure that larger disks are
supported on ATA100.

Since the 8KHA+ is a fairly old board, there is a good chance that it
is not supported.

I'm suprised at the terrible answer from Epox. They never tested with
larger then 80GB? They don't know the specifications of their own
boards?
The person at Epox that gave that answer should be ashamed of himself.

Marc

"We have tested up to 80GB on this model."

I got exactly that one line answer. It wasn't even slightly useful since I
already have an 80 gig drive, so I know that it works.

You would think they wouldn't have to test, right? Aside from bugs, shouldn't
they know how much it will support simply because they have to *design* it to
support a certain size?

But I guess tech support card readers wouldn't know anything about that.

Thanks for the detailed info.

 

[Eric Gisin]

Bullshit. LBA-48 was introduced in ATA-6, which doesn't even mention
UDMA/133. Anyone know which came to market first?

You are far better off looking at your BIOS date rather than UDMA speeds.

 

[Rod Speed]

That just depends on the age of your mobo.

Nope, what the bios can do.

He is right that initially ATA100 boards were limited to 137MB. One
of the improvements of ATA133 was the support for larger disks.

Utterly mangled. There are plenty of ATA100
systems that support those bigger drives.

But newer ATA100 devices also support larger disks now.

So he was just plain wrong, like I said.

(Lots of companies didn't want to use ATA133 as defined
by Maxtrox? But they did see the need for larger disks)

48bit LBA is an entirely separate issue to ATA133.

So the correct answer is that you are sure that larger
disks are supported on ATA133, but you cannot be
sure that larger disks are supported on ATA100.

Its more complicated than that too. Plenty of
systems that claim to support say ATA133 dont
have everything its supposed to have implemented.

Since the 8KHA+ is a fairly old board, there
is a good chance that it is not supported.

I didnt even comment on that at all, JUST on his much
too simplistic and just plain wrong statements quoted.

I'm suprised at the terrible answer from Epox.

Yeah, looks like either the wires got crossed
with someone not having a very good grasp of
english or some pig ignorant monkey involved.

They never tested with larger then 80GB?

Corse they have. Presumably they actually mean that
that particular older motherboard was never CERTIFIED
with drives bigger than that, a different matter entirely.

They don't know the specifications of their own boards?
The person at Epox that gave that answer should be ashamed of himself.

Yeah, tho we dont know the exact words used in the question.

 

[Folkert Rienstra]

That just depends on the age of your mobo.

He is right that initially ATA100 boards were limited to 137MB.

One of the improvements of ATA133 was the support for larger disks.

No, it was not.

But newer ATA100 devices also support larger disks now.

Of course they do. It's a software thing, not hardware.

(Lots of companies didn't want to use ATA133 as defined by Maxtrox?
Who?

But they did see the need for larger disks)

So the correct answer is that you are sure that larger disks are
supported on ATA133, but you cannot be sure that larger disks
are supported on ATA100.

Which entirely depends on the manufacturer of the chipset
or the maintainer of the OS's willingness to support.

Since the 8KHA+ is a fairly old board, there is a good chance that
it is not supported.

I'm suprised at the terrible answer from Epox.

Al least it was an honest answer if somewhat shortsighted.

They never tested with larger then 80GB? They don't know the
specifications of their own boards?

When a third party supplies your bios, that may happen in the absence
of still to come out drive capacities.

The person at Epox that gave that answer should be ashamed of himself.

Well, you're not that much better.

 

[Marc de Vries]

Nope, what the bios can do.

which is depended upon the availability of updates, which is in
practice depended on the age of your mobo.

Utterly mangled. There are plenty of ATA100
systems that support those bigger drives.

So my reply to Folkert.
It seems those ATA100 systems are also ATA-6 which does support bigger
drives.

So he was just plain wrong, like I said.

Most of what he said was correct. ATA/33 and ATA/66 doesn't support 48
bit addressing. Some ATA/100 systems don't either.

ATA/100 systems that are ATA-6 support larger disks
ATA/133 systems all seem to support 48 bit addressing.

There is just some confusion about ATA100

48bit LBA is an entirely separate issue to ATA133.

Correct.
But Maxtor has done it's best to let people think that it was not a
seperate issue.

Its more complicated than that too.

Yeah I noticed that. This whole ATA business is one big mess.

Plenty of
systems that claim to support say ATA133 dont
have everything its supposed to have implemented.

Actually it seems like the only thing that you need to support for
ATA133 is 133MB/s.

I haven't found documents which state that ATA133 also supports other
features. Those would probably be defined in ATA-6 or ATA-7

Yeah, looks like either the wires got crossed
with someone not having a very good grasp of
english or some pig ignorant monkey involved.

It's a supportdesk. I'd bet on the last option
(just had some bad experience this week with such support desks...)

Corse they have. Presumably they actually mean that
that particular older motherboard was never CERTIFIED
with drives bigger than that, a different matter entirely.

Hmm. You are probably right about that. I didn't think of that option.
The OP should probably have asked them not to state the certified
size, but the theoretical.

Marc

 

[Marc de Vries]

No, it was not.

Oh, it was just a coincidence that the first disks larger than 137GB
had ATA133?

Well, I decided to do some digging into this matter and it wasn't a
conicidence. Here is what I found at storagereview:

"Perhaps to make ATA-133 more attractive, Maxtor has bundled the "Big
Drive" benefits of 48-bit addressing with ATA-133 in both marketing
and in the physical controller bundled with the 160 GB units. To
overcome the inherent capacity limitations present in most of today's
ATA controllers, a Maxtor-branded Promise Ultra133TX2 controller comes
bundled in each box. Promise's card not only offers ATA-133 operation
but also includes the aforementioned 48-bit addressing.."
http://www.storagereview.com/articles/200112/200112204G160J8_1.html

So the confusion is certainly not strange and in fact caused
intentionally by Maxtor.


But basically I found that the AT standard is a big mess.

Hardware manufacturers, reviews sites and software manufacturers often
don't stick to the correct terms or even mix the terms.

The technical documentation of the IBM 75GXP only specifies that the
interface standard is ATA/100, but no mention is made at all whether
it is also ATA-6
But the technical documentation of the IBM IBM 180GXP specifies at the
same place in the documentation that the interface standard is ATA-6
(The disk is also ATA/100 but this time no further mention of this)

Anandtech give the following information about ata:
* ATA-5 (ATA100, Ultra DMA100, DMA100) - Supports DMA mode 5. Runs at
a maximum of 100MBps.
* ATA-6 (ATA133, Ultra DMA133, DMA133) - Supports DMA mode 6. Runs at
a maximum of 133MBps. Adds support for drives larger then 137GB

But Tom's Hardware guide states
* ATA-5 Ultra ATA/66
* ATA-6 Ultra ATA/100 and 48 bit

Others sites again list:
Ultra DMA 4 as ATA/66
Ultra DMA 5 as ATA/100 (which probably explains the mistake by
Anandtech)
Ultra DM 6 as ATA/133


From http://ata-atapi.com/hist.htm#T20 comes:

ATA/ATAPI-5 deletes a few old commands, adds a few new commands,
changes the way a few commands operate. But the big thing in
ATA/ATAPI-5 are the two new and faster Ultra DMA 66 data transfer
modes.

ATA/ATAPI-6 includes another even faster Ultra DMA mode 5, also known
as Ultra DMA 100. It also includes a method of increasing the number
of LBA bits from 28 to 48 and increasing the Sector Count from 8 bits
to 16 bits.

Work on ATA/ATAPI-6 was completed at the T13 meeting in October 2001.
ATA/ATAPI-6 should be a published ANSI standard in early 2002.

But the IBM 75GXP was already in production in august 2000.

So what standards did the 75GXP follow? It uses ATA/100 which is not
supported by ATA-5, but was completed before the ATA-6 standard was
available.

But in june 2001 Maxtor had already started their bigdrive initiative
wich added 48 bit addressing.
In september 2001 (so 1 month after work on ATA/6 with ATA/100 was
complete) Maxtor pushed their 160BG disks with ATA/133 and pretended
that 48 bit addressing was something of the ATA/133 standard.
Maybe even technically correct, because at the time of their statement
the ATA-6 standard might not have been completed.

You are correct that ATA133 is seperate from 48 bit addressing. (Or
big drive initiative as Maxtor likes to call it)


It seems that ATA100 and ATA133 are even completely seperate from the
ATA-6 standard.
The document on the ATA-6 standard doesn't make any mention at all
about 100MB/s or 133MB/s: http://www.t13.org/project/d1410r0.pdf

Some sites claim that ATA-6 supports both 100MB/s and 133MB/s, other
claim that 133MB/s is part of ATA-7

Unfortunately most hardware manufacturers and websites don't make any
distinction between AT6 and ATA100 or ATA133.

So if a companies states that your device (for instance a raid
controller) is ATA100 you can't say anything about 48 bit addressing.
If it states ATA133, then theoretically you can't say anything about
48bit addressing either, but because Maxtor has pushed ATA133 and
48bit addressing as one big feature, you can be 99% sure that 48 bit
addressing is also supported.

Of course they do. It's a software thing, not hardware.

I didn't say newer ATA100 harddisks. I said devices. This also
includes ATA/100 IDE controllers which DO need support for this in
their firmware. Which many would consider hardware.
(although you might want to call this software too, if you think this
whole business isn't confusing enough already)

Older ATA/100 IDE controllers and ATA/100 IDE Raid controllers do not
support this out of the box.

Who?

IBM/Hitachi for instance. Even their latest harddisks are all ATA 100.

Which entirely depends on the manufacturer of the chipset
or the maintainer of the OS's willingness to support.

True. but that does not invalidate my statement.

Al least it was an honest answer if somewhat shortsighted.


When a third party supplies your bios, that may happen in the absence
of still to come out drive capacities.

Even then they should know the specifications of bios supplied by
their that third party, which should list whether or not it has 48 bit
addressing/

Well, you're not that much better.

First of all it is not my job to know those facts. I'm not paid to do
this.
I haven't seen any payment from you. You can give such criticism once
I have received your payment.

Second: You have not invalidated my statement that you cannot be sure
with ATA/100 devices, but can be sure with ATA/133 devices. The
statement was correct, even though ATA/133 is technically not directly
related to 48 bit addressing.

Third: You have not made the distinction between ATA 100/133 and ATA6
ATA7 anymore then I did. So if I should be ashamed, then you should be
ashamed too!!

Marc

 

[JustDoIt]

It's nice to see how much info you can get when people get offended
I had no clue of all this ATA33/66/100/133/4/5/6/7/?, allways looked at
33/66/100/133/?
TNX all

 

[Rod Speed]

which is depended upon the availability of updates,

Nope, a bios overlay can be used instead.

which is in practice depended on the age of your mobo.

Wrong again. You arent even necessarily dependant on the
motherboard manufacturer on that and they vary a hell of a
lot on what they do about that sort of support for older bios.

So my reply to Folkert.

Completely useless. 48bit LBA support has
NOTHING to do with the ATAxxx number at all.

It seems those ATA100 systems are also
ATA-6 which does support bigger drives.

And 48bit LBA support isnt only seen with ATA100 and better.

Most of what he said was correct.

Like hell it was. 48bit LBA support has NOTHING to
do with what speed is supported on the ribbon cable.

ATA/33 and ATA/66 doesn't support 48 bit addressing.

Wrong again.

Some ATA/100 systems don't either.

So his original is just plain wrong.

ATA/100 systems that are ATA-6 support larger disks

ATA100 system may or may not have 48bit LBA support.

ATA/133 systems all seem to support 48 bit addressing.

Wrong again.

There is just some confusion about ATA100

Wrong again.

AND the ATAxxx doesnt specify exactly what is supported
and what isnt. JUST the best speed possible.

Correct.

So his original is just plain wrong.

But Maxtor has done it's best to let people
think that it was not a seperate issue.

Completely irrelevant to whether its an entirely separate issue or not.

That may well be what produced his original statement, but its
always been wrong and is still wrong. 48bit LBA support is an
entirely separate issue to ATAxxx number and can be added to
a bios of a motherboard that only supports ATA66 if you want to.

Some have done just that.

Yeah I noticed that. This whole ATA business is one big mess.

Yeah, its better than it used to be but still one hell of a mess,
particularly with stuff like 48bit LBA support which is an entirely
separate issue to the maximum speed supported on the ribbon cable.

And there is no such animal as ATA133 standards wise anyway.

Actually it seems like the only thing that
you need to support for ATA133 is 133MB/s.

Precisely. Thats waying what I have been saying all along in different words.
That maximum speed supported is an entirely separate issue to whether
there is 48bit LBA support and 48bit LBA support can be added to a bios
for a motherboard that only supports a max of 66 too if you want to.

I haven't found documents which state
that ATA133 also supports other features.

Basically because ATA133 is just a specification
the maximum speed supported and nothing else.

Those would probably be defined in ATA-6 or ATA-7

Not in the sense that there is a rigid requirement that a system
that can only do say 66 or 100 cant have 48bit LBA too.

It's a supportdesk. I'd bet on the last option

Dunno, one massive problem with the type of operation being
discussed is that english is only their second language at best
and that causes quite a few problems. It wouldnt be at all
surprising if the individual that replied by email assumed that
what was being asked was what that particular motherboard
had been CERTIFIED to be able to do rather than what was
actually being asked, what should would fine, but which hasnt
actually been CERTIFIED by the manufacturer, just because
its a well obsolete board now that they no longer bother to
certify anymore.

(just had some bad experience this week with such support desks...)

Yeah, anyone with a clue doesnt need to
'work' in such a hopeless dead end 'job'

Hmm. You are probably right about that. I didn't think
of that option. The OP should probably have asked
them not to state the certified size, but the theoretical.

Yeah, thats what I meant about poor english.
Likely thats how the problem arose.

 

[Rod Speed]

Oh, it was just a coincidence that the first
disks larger than 137GB had ATA133?

Yep. Maxtor started flogging drives with ATA133 purely because
that looked better, at a time when the drives didnt need that.

They also were one of the first to flog drives over 137GB.

Well, I decided to do some digging into this matter and it
wasn't a conicidence. Here is what I found at storagereview:

"Perhaps to make ATA-133 more attractive, Maxtor has bundled
the "Big Drive" benefits of 48-bit addressing with ATA-133 in both
marketing and in the physical controller bundled with the 160 GB units.

Thats rather silly. The real reason for bundling the controller
with the drive was just to simplify installation of the drive in
systems which may not have 48bit LBA support.

To overcome the inherent capacity limitations
present in most of today's ATA controllers,

Most bios, actually.

a Maxtor-branded Promise Ultra133TX2 controller comes
bundled in each box. Promise's card not only offers ATA-133
operation but also includes the aforementioned 48-bit addressing.."
http://www.storagereview.com/articles/200112/200112204G160J8_1.html

So the confusion is certainly not strange
and in fact caused intentionally by Maxtor.

Nope. That ease of installation is an entirely separate issue again.

But basically I found that the AT standard is a big mess.

Sure, in the sense that there isnt a nice tidy list of what
must be included to be able to claim ATA100 or ATA133.

That maximum speed is just ONE feature, not the entire standard.

Hardware manufacturers, reviews sites and software manufacturers
often don't stick to the correct terms or even mix the terms.

Yep. Because ATA-6 for example includes a lot of different
features, quite a few of which are optional and have nothing
to do with the maximum speed the implimentation supports.

The technical documentation of the IBM 75GXP only
specifies that the interface standard is ATA/100, but
no mention is made at all whether it is also ATA-6

Yep, because even ATA-6 isnt a nice tidy list that specifys what
is certain to be included with a system that has a 100 max speed.

But the technical documentation of the IBM IBM 180GXP specifies at the
same place in the documentation that the interface standard is ATA-6

Which still has a great raft of optional stuff. And 48bit LBA support
isnt even a feature of a drive anyway, its at the other end of the cable.

(The disk is also ATA/100 but this time no further mention of this)

Anandtech give the following information about ata:
* ATA-5 (ATA100, Ultra DMA100, DMA100) -
Supports DMA mode 5. Runs at a maximum of 100MBps.

* ATA-6 (ATA133, Ultra DMA133, DMA133) - Supports DMA mode 6.
Runs at a maximum of 133MBps. Adds support for drives larger then 137GB

But doesnt tie 48bit LBA support to the maximum speed supported.

But Tom's Hardware guide states
* ATA-5 Ultra ATA/66
* ATA-6 Ultra ATA/100 and 48 bit

Others sites again list:
Ultra DMA 4 as ATA/66
Ultra DMA 5 as ATA/100 (which probably
explains the mistake by Anandtech)
Ultra DM 6 as ATA/133

From http://ata-atapi.com/hist.htm#T20 comes:

ATA/ATAPI-5 deletes a few old commands, adds a few new commands,
changes the way a few commands operate. But the big thing in
ATA/ATAPI-5 are the two new and faster Ultra DMA 66 data transfer
modes.

ATA/ATAPI-6 includes another even faster Ultra DMA mode 5,
also known as Ultra DMA 100. It also includes a method of
increasing the number of LBA bits from 28 to 48 and
increasing the Sector Count from 8 bits to 16 bits.

But does NOT tied 48bit LBA support to the maximum speed possible.

Those are entirely separate features.

Work on ATA/ATAPI-6 was completed at the T13 meeting in October 2001.
ATA/ATAPI-6 should be a published ANSI standard in early 2002.

But the IBM 75GXP was already in production in august 2000.

So what standards did the 75GXP follow?

Irrelevant when 48bit LBA support isnt even in the drive.

It uses ATA/100 which is not supported by ATA-5, but
was completed before the ATA-6 standard was available.

Yep, just another example of where the
ATA standard lags significantly at times.

But in june 2001 Maxtor had already started their
bigdrive initiative wich added 48 bit addressing.

It wasnt a Maxtor initiative as you imply there.

In september 2001 (so 1 month after work on
ATA/6 with ATA/100 was complete) Maxtor pushed
their 160BG disks with ATA/133 and pretended that
48 bit addressing was something of the ATA/133 standard.

There is no such animal as 'the ATA/133 standard'. ATA133 is just
an informal way of specifying the maximum speed supported without
even necessarily specifying exactly which DMA modes are supported.

And the formal approval dates are largely irrelevant anyway
when quite a bit of the time its obvious well ahead of the formal
completion date whats going to end up in the new revision.

Maybe even technically correct, because at the time of their
statement the ATA-6 standard might not have been completed.

You are correct that ATA133 is seperate from 48 bit addressing.

Which is all I ever said all along.

(Or big drive initiative as Maxtor likes to call it)

Just the usual manufacturer wanking.

It seems that ATA100 and ATA133 are even
completely seperate from the ATA-6 standard.

Yep, they have never been a proper formal anything, just
an informal way of specifying the maximum speed possible.

The document on the ATA-6 standard doesn't make any mention at all
about 100MB/s or 133MB/s: http://www.t13.org/project/d1410r0.pdf

Some sites claim that ATA-6 supports both 100MB/s and
133MB/s, other claim that 133MB/s is part of ATA-7

Just irrelevant hair splitting as far as 48bit
LBA support being discussed is concerned.

Unfortunately most hardware manufacturers and websites don't
make any distinction between AT6 and ATA100 or ATA133.

Yep, what matters with a drive is the maximum speed supported
and even 133 is an irrelevant wank for many drives that support it.
Just makes the drive look better to the pig ignorant consumer.

So if a companies states that your device (for instance
a raid controller) is ATA100 you can't say anything about
48 bit addressing. If it states ATA133, then theoretically
you can't say anything about 48bit addressing either,

Which is what I said all along.

It doesnt even say anything about 48bit LBA support with
a motherboard that only supports a max of ATA66 either.

All you can really say is that the later the device, the more
LIKELY it is to have 48bit LBA support, which is hardly
a shocking revelation or any news to anyone with a clue.

but because Maxtor has pushed ATA133 and
48bit addressing as one big feature, you can be
99% sure that 48 bit addressing is also supported.

Its completely irrelevant what Maxtor is pushing.

What matters is that there are plenty of mass market
160GB drives around now, which need 48bit LBA
support. In fact everyone has drives that big now.

I didn't say newer ATA100 harddisks. I said devices.

He was obviously talking about devices because 48bit LBA
is a software thing in the device the hard drive is attached to.

This also includes ATA/100 IDE controllers
which DO need support for this in their firmware.
Which many would consider hardware.

Only the terminally pig ignorant.

(although you might want to call this software too,

Corse its software.

if you think this whole business isn't confusing enough already)

Older ATA/100 IDE controllers and ATA/100 IDE
Raid controllers do not support this out of the box.

Older anything doesnt. So what ?

Plenty of motherboards with a max of ATA100 do support 48bit LBA.

IBM/Hitachi for instance. Even their latest harddisks are all ATA 100.

He meant who Matrox ? You obviously meant Maxtor.

True. but that does not invalidate my statement.

Corse it does. Your certainty is just plain wrong.

Even then they should know the specifications
of bios supplied by their that third party, which
should list whether or not it has 48 bit addressing.

Not necessarily at the time it was used.

First of all it is not my job to know those facts. I'm not paid
to do this. I haven't seen any payment from you. You can
give such criticism once I have received your payment.

All completely irrelevant to whether you did much better.

Second: You have not invalidated my statement
that you cannot be sure with ATA/100 devices,
but can be sure with ATA/133 devices.

Its so obviously wrong that that isnt necessary.

The statement was correct,
Nope.

even though ATA/133 is technically not
directly related to 48 bit addressing.

It isnt practically either in the way you claim, particularly with
motherboards, which just happens to what was being discussed.

Third: You have not made the distinction between
ATA 100/133 and ATA6 ATA7 anymore then I did.

Not relevant to what was being discussed.

So if I should be ashamed, then you should be ashamed too!!

Bet that will make him shrivel |-)

 

[Folkert Rienstra]

Oh, it was just a coincidence that the first disks larger than 137GB
had ATA133?
Yes.


Well, I decided to do some digging into this matter and it wasn't a
conicidence.

Here is what I found at storagereview:

Oh wow, then it must be true.

"Perhaps

Now there's a good start.

to make ATA-133 more attractive, Maxtor has bundled the "Big
Drive" benefits of 48-bit addressing with ATA-133 in both marketing
and in the physical controller bundled with the 160 GB units. To
overcome the inherent capacity limitations present in most of today's
ATA controllers, a Maxtor-branded Promise Ultra133TX2 controller comes
bundled in each box. Promise's card not only offers ATA-133 operation
but also includes the aforementioned 48-bit addressing.."
http://www.storagereview.com/articles/200112/200112204G160J8_1.html

So the confusion is certainly not strange and in fact caused intentionally by Maxtor.

Nope, Storage Review, obviously.

But basically I found that the AT standard is a big mess.

Hardware manufacturers, reviews sites and software manufacturers
often don't stick to the correct terms or even mix the terms.

Thanks for undermining your own comments above so directly.

The technical documentation of the IBM 75GXP only specifies that the
interface standard is ATA/100, but no mention is made at all whether
it is also ATA-6

It isn't. It probably says "Enhanced IDE (ATA-5) interface".
Anyway, that's what it says for the 60GXP, which is newer.

But the technical documentation of the IBM IBM 180GXP specifies at the
same place in the documentation that the interface standard is ATA-6
(The disk is also ATA/100 but this time no further mention of this)

Anandtech give the following information about ata:
* ATA-5 (ATA100, Ultra DMA100, DMA100) - Supports DMA mode 5.
Runs at a maximum of 100MBps.
* ATA-6 (ATA133, Ultra DMA133, DMA133) - Supports DMA mode 6.
Runs at a maximum of 133MBps. Adds support for drives larger then 137GB

But Tom's Hardware guide states
* ATA-5 Ultra ATA/66
* ATA-6 Ultra ATA/100 and 48 bit

Others sites again list:
Ultra DMA 4 as ATA/66
Ultra DMA 5 as ATA/100 (which probably explains the mistake by Anandtech)

What mistake?

Ultra DMA-6 as ATA/133


From http://ata-atapi.com/hist.htm#T20 comes:

ATA/ATAPI-5 deletes a few old commands, adds a few new commands,
changes the way a few commands operate. But the big thing in ATA/
ATAPI-5 are the two new and faster Ultra DMA 66 data transfer modes.

ATA/ATAPI-6 includes another even faster Ultra DMA mode 5, also known
as Ultra DMA 100. It also includes a method of increasing the number of LBA
bits from 28 to 48 and increasing the Sector Count from 8 bits to 16 bits.

Work on ATA/ATAPI-6 was completed at the T13 meeting in October 2001.
ATA/ATAPI-6 should be a published ANSI standard in early 2002.

But the IBM 75GXP was already in production in august 2000.

So what standards did the 75GXP follow? It uses ATA/100 which is not
supported by ATA-5,

Says who?

but was completed before the ATA-6 standard was available.

But in june 2001 Maxtor had already started their bigdrive initiative
wich added 48 bit addressing.
In september 2001 (so 1 month after work on ATA/6 with ATA/100 was
complete) Maxtor pushed their 160BG disks with ATA/133 and pretended
that 48 bit addressing was something of the ATA/133 standard.
Maybe even technically correct, because at the time of their statement
the ATA-6 standard might not have been completed.

You are correct that ATA133 is seperate from 48 bit addressing. (Or
big drive initiative as Maxtor likes to call it)


It seems that ATA100 and ATA133 are even completely seperate from the
ATA-6 standard.
The document on the ATA-6 standard doesn't make any mention at all
about 100MB/s or 133MB/s: http://www.t13.org/project/d1410r0.pdf

An omission in annex C: Signal integrity and UDMA implementation guide
of d1410r3.pdf May well have been taken care of in d1410r3b.pdf

Some sites claim that ATA-6 supports both 100MB/s and 133MB/s, other
claim that 133MB/s is part of ATA-7

Unfortunately most hardware manufacturers and websites don't make any
distinction between ATA-6 and ATA100 or ATA133.

So if a companies states that your device (for instance a raid
controller) is ATA100 you can't say anything about 48 bit addressing.
If it states ATA133, then theoretically you can't say anything about
48bit addressing either, but because Maxtor has pushed ATA133 and
48bit addressing as one big feature, you can be 99% sure that 48 bit
addressing is also supported.


I didn't say newer ATA100 harddisks.

Nor did I.

I said devices. This also includes ATA/100 IDE controllers which DO
need support for this in their firmware.

Nope, just bios and driver.

Which many would consider hardware.

Nope, just BIOS and driver, i.e. software.

(although you might want to call this software too, if you think this
whole business isn't confusing enough already)

Older ATA/100 IDE controllers and ATA/100 IDE Raid controllers do not
support this out of the box.

As said before, this is a matter of (bios and driver) support.

Maxtrox?


IBM/Hitachi for instance. Even their latest harddisks are all ATA 100.


True. but that does not invalidate my statement.

Actually it does when the ATA133 controller came out before the big
drives really came in to being and the chip then got followed up by a
newer one (64-bit PCI) and the older controller is no longer supported.

Even then they should know the specifications of bios supplied by
their that third party, which should list whether or not it has 48-bit
addressing/

That doesn't say anything about the bugs in the bios.

First of all it is not my job to know those facts. I'm not paid to do this.

Ah, so it is fine then to lie or speak nonsense when you are not paid for it.

I haven't seen any payment from you. You can give such criticism once
I have received your payment.

Second: You have not invalidated my statement that you cannot be sure
with ATA/100 devices, but can be sure with ATA/133 devices.

Actually, you did that yourself just fine, a few lines ago.

The statement was correct, even though ATA/133 is technically not directly
related to 48 bit addressing.

Third: You have not made the distinction between ATA 100/133 and ATA-6
ATA-7 anymore than I did.

I didn't make the initial statement, you did.

 

[Folkert Rienstra]

which is depended upon the availability of updates, which is in
practice depended on the age of your mobo.

Now you've got it.

See my reply to Folkert.
It seems those ATA100 systems are also ATA-6 which does support
bigger drives.

It's a software matter so should be ATA spec independent on the
host controller hardware side.

Most of what he said was correct. ATA/33 and ATA/66 doesn't
support 48-bit addressing. Some ATA/100 systems don't either.

Because of lacking support.

ATA/100 systems that are ATA-6 support larger disks
ATA/133 systems all seem to support 48 bit addressing.

A sound commercial choice when you want to sell new kit (make profit).

There is just some confusion about ATA100


Correct.
But Maxtor has done it's best to let people think that it was not a
seperate issue.

Oh? I got a very different impression.
IIRC, they divided the issue in Fast drives and Big drives.

Yeah I noticed that. This whole ATA business is one big mess.


Actually it seems like the only thing that you need to support for
ATA133 is 133MB/s.

I haven't found documents which state that ATA133 also supports other
features.

Of course not, as it is the spec that defines UDMA
mode 6 and not UDMA mode 6 describing the spec.