eSata drivers for internal eSata drives under WinXP SP3

Does Microsoft have available for download drivers which will improve the
performance of eSata internal harddrives.
As shipped originally WinXP was probably not equipped to handle eSata, but
years later there are many things available for use under WinXP SP3 which
the old drivers won't optimize.

--
Kindly remove my name and address before forwarding this e-mail. Also,
please place my address in the bcc field when sending me e-mail.
We have no control over who will see forwarded messages! This helps to keep
all our Personal Contacts lists Private and Stops Intruders & Spammers.

From: "Docster"

Does Microsoft have available for download drivers which will improve the
performance of eSata internal harddrives.
As shipped originally WinXP was probably not equipped to handle eSata, but
years later there are many things available for use under WinXP SP3 which
the old drivers won't optimize.


By definition eSATA are Externally connectected SATA Drives not "eSata
internal harddrives".

Based upon a given system, there are internal and external connecttions but
there is only one OS SATA driver for the SATA controller and associated
chip-set.

The concept is OS independent.


--
Dave
Multi-AV Scanning Tool - http://multi-av.thespykiller.co.uk
http://www.pctipp.ch/downloads/dl/35905.asp

Docster wrote:
Does Microsoft have available for download drivers which will improve the
performance of eSata internal harddrives.
As shipped originally WinXP was probably not equipped to handle eSata, but
years later there are many things available for use under WinXP SP3 which
the old drivers won't optimize.


ESATA is the same thing as SATA.

How they differ, is the electrical levels are slightly different.

http://en.wikipedia.org/wiki/Serial_ATA#eSATA

"transmit amplitude... 500–600 mV"
"receive amplitude... 240–600 mV"

The purpose of doing that, is to support a 2 meter cable,
instead of a 1 meter cable. The difference between launch and
receive sensitivity, gives room for the loss caused by a longer cable.

While an Intel spec happens to repeat that info, there is nothing
that I could find, to suggest a port was programmable as SATA or
ESATA (i.e. change the receiver or transmitter characteristics).
I have to conclude from that, the port runs ESATA levels at all
times, even when regular SATA drives are connected. I don't
think the port is in a position to "guess" what is connected.
And unless I can find a register that sets those levels,
then I have to conclude it uses ESATA levels, so it is ready
for ESATA when it shows up.

*******

What WinXP is not equipped with, is AHCI drivers. You have to use
a manufacturer AHCI driver for that. And it's debatable whether
that gives an actual performance improvement. The workload on a
desktop, probably doesn't build enough queue depth, for native
command queuing to gain an advantage. And while the hot plug
feature of AHCI is fun, it has to be done with some care, so
the drive doesn't get damaged in handling (bump the drive while
the platter is spinning).

http://en.wikipedia.org/wiki/Ahci

"Many SATA controllers offer selectable modes of operation:
legacy Parallel ATA emulation,
standard AHCI mode
vendor-specific RAID"

WinXP has two drivers available for legacy parallel ATA emulation.
One driver handles compatible mode (I/O space, INT14 and INT15),
while the other driver handles PCI space, INTA#. In cases where
a chip happened to not support that emulation, then a vendor
specific driver would be needed.

*******

What will improve the performance of SATA/ESATA devices,
would be a caching controller, where the OS doesn't know
caching is occurring.

But that's not a practical solution. Using an SSD for
the boot drive, is the most practical. At least one
poster to USENET, has switched entirely to SSD, and
owns multiple of them. Whereas I own zero of them

Paul

"Paul" wrote in message
...
Docster wrote:
Does Microsoft have available for download drivers which will improve the
performance of eSata internal harddrives.
As shipped originally WinXP was probably not equipped to handle eSata,
but years later there are many things available for use under WinXP SP3
which the old drivers won't optimize.


ESATA is the same thing as SATA.

How they differ, is the electrical levels are slightly different.

http://en.wikipedia.org/wiki/Serial_ATA#eSATA

"transmit amplitude... 500–600 mV"
"receive amplitude... 240–600 mV"

The purpose of doing that, is to support a 2 meter cable,
instead of a 1 meter cable. The difference between launch and
receive sensitivity, gives room for the loss caused by a longer cable.

While an Intel spec happens to repeat that info, there is nothing
that I could find, to suggest a port was programmable as SATA or
ESATA (i.e. change the receiver or transmitter characteristics).
I have to conclude from that, the port runs ESATA levels at all
times, even when regular SATA drives are connected. I don't
think the port is in a position to "guess" what is connected.
And unless I can find a register that sets those levels,
then I have to conclude it uses ESATA levels, so it is ready
for ESATA when it shows up.

*******

What WinXP is not equipped with, is AHCI drivers. You have to use
a manufacturer AHCI driver for that. And it's debatable whether
that gives an actual performance improvement. The workload on a
desktop, probably doesn't build enough queue depth, for native
command queuing to gain an advantage. And while the hot plug
feature of AHCI is fun, it has to be done with some care, so
the drive doesn't get damaged in handling (bump the drive while
the platter is spinning).

http://en.wikipedia.org/wiki/Ahci

"Many SATA controllers offer selectable modes of operation:
legacy Parallel ATA emulation,
standard AHCI mode
vendor-specific RAID"

WinXP has two drivers available for legacy parallel ATA emulation.
One driver handles compatible mode (I/O space, INT14 and INT15),
while the other driver handles PCI space, INTA#. In cases where
a chip happened to not support that emulation, then a vendor
specific driver would be needed.

*******

What will improve the performance of SATA/ESATA devices,
would be a caching controller, where the OS doesn't know
caching is occurring.

But that's not a practical solution. Using an SSD for
the boot drive, is the most practical. At least one
poster to USENET, has switched entirely to SSD, and
owns multiple of them. Whereas I own zero of them

Paul

I bought a Kingston 128gb SSD kit for my MSI CR620 and it was initially
lovely. Rummaged around, made sure that TRIM was working etc. And just
after fitting it, as a precaution, I spashed out on Acronis TrueImage and
imaged the 128gb SSD onto the 256gb hd now in the external usb caddy
supplied with the kit,

......after a couple of weeks Laptop refused to boot !!! ...so had to
restore hard disk partitions from external 256gb 5,600 rpm hd, to SSD in
Laptop.

I suppose I could have left original 256gb 5,600rpm? hd as it was and then
refitted it to Laptop, and gone through the whole procedure again but,
having used Acronis to image the SSD as a precaution, ....that turned out to
be a a wise and time saving precuation !!!

So I have no idea why SSD suddenly refused to boot, after being in there a
couple of weeks, and since restoring SSD image external 256gb hd, it's
performed flawlesslly for over 6 months, and never refuses to boot !!!

regards, Richard

I would never advise anyone to swap to a SSD unless, IMMEDIATELY after
fitting it, one imaged the thing onto another drive, and kept that image up
to date.

RJK wrote:
"Paul" wrote in message
...
Docster wrote:
Does Microsoft have available for download drivers which will improve the
performance of eSata internal harddrives.
As shipped originally WinXP was probably not equipped to handle eSata,
but years later there are many things available for use under WinXP SP3
which the old drivers won't optimize.

ESATA is the same thing as SATA.

How they differ, is the electrical levels are slightly different.

http://en.wikipedia.org/wiki/Serial_ATA#eSATA

"transmit amplitude... 500–600 mV"
"receive amplitude... 240–600 mV"

The purpose of doing that, is to support a 2 meter cable,
instead of a 1 meter cable. The difference between launch and
receive sensitivity, gives room for the loss caused by a longer cable.

While an Intel spec happens to repeat that info, there is nothing
that I could find, to suggest a port was programmable as SATA or
ESATA (i.e. change the receiver or transmitter characteristics).
I have to conclude from that, the port runs ESATA levels at all
times, even when regular SATA drives are connected. I don't
think the port is in a position to "guess" what is connected.
And unless I can find a register that sets those levels,
then I have to conclude it uses ESATA levels, so it is ready
for ESATA when it shows up.

*******

What WinXP is not equipped with, is AHCI drivers. You have to use
a manufacturer AHCI driver for that. And it's debatable whether
that gives an actual performance improvement. The workload on a
desktop, probably doesn't build enough queue depth, for native
command queuing to gain an advantage. And while the hot plug
feature of AHCI is fun, it has to be done with some care, so
the drive doesn't get damaged in handling (bump the drive while
the platter is spinning).

http://en.wikipedia.org/wiki/Ahci

"Many SATA controllers offer selectable modes of operation:
legacy Parallel ATA emulation,
standard AHCI mode
vendor-specific RAID"

WinXP has two drivers available for legacy parallel ATA emulation.
One driver handles compatible mode (I/O space, INT14 and INT15),
while the other driver handles PCI space, INTA#. In cases where
a chip happened to not support that emulation, then a vendor
specific driver would be needed.

*******

What will improve the performance of SATA/ESATA devices,
would be a caching controller, where the OS doesn't know
caching is occurring.

But that's not a practical solution. Using an SSD for
the boot drive, is the most practical. At least one
poster to USENET, has switched entirely to SSD, and
owns multiple of them. Whereas I own zero of them

Paul

I bought a Kingston 128gb SSD kit for my MSI CR620 and it was initially
lovely. Rummaged around, made sure that TRIM was working etc. And just
after fitting it, as a precaution, I spashed out on Acronis TrueImage and
imaged the 128gb SSD onto the 256gb hd now in the external usb caddy
supplied with the kit,

.....after a couple of weeks Laptop refused to boot !!! ...so had to
restore hard disk partitions from external 256gb 5,600 rpm hd, to SSD in
Laptop.

I suppose I could have left original 256gb 5,600rpm? hd as it was and then
refitted it to Laptop, and gone through the whole procedure again but,
having used Acronis to image the SSD as a precaution, ....that turned out to
be a a wise and time saving precuation !!!

So I have no idea why SSD suddenly refused to boot, after being in there a
couple of weeks, and since restoring SSD image external 256gb hd, it's
performed flawlesslly for over 6 months, and never refuses to boot !!!

regards, Richard

I would never advise anyone to swap to a SSD unless, IMMEDIATELY after
fitting it, one imaged the thing onto another drive, and kept that image up
to date.


That seems to be the failure pattern.

The SSD is more likely to fail suddenly. And not in response
to "wear out". A typical failure mechanism is firmware related.
The firmware version inside the SSD has a problem. When you buy
an SSD, the first thing you do, is investigate how many
firmwares were released, and what problems they have.

Your symptoms are a bit different, in that the device still functions.
Now, maybe there was a bad block in the flash, and that bad block
has been spared out (taking one of your boot files with it). The
drive has a certain amount of redundancy, and spare flash blocks
are available. Maybe that is why it works today.

In any case, "do frequent backups". That's my advice for SSD.
Some brands seem to be pretty well designed, but there
are a fair number of "sad stories" out there, of data loss.

On a hard drive, if you check the SMART statistics once in
a while, you can get a warning that the drive is developing
problems. And perhaps, before they become visible in normal
operation. While SSDs have SMART (the parameters aren't
the same set as with a hard drive), I don't think there is
any parameter that warns about "sudden failure" type problems.

Paul