Windows XP SSD tweaks?

[John Doe]

....

This article mentions turning off "timestamp for last access".

http://www.ocztechnologyforum.com/forum/showthread.php?t=43460

From the same forum... I tried aligning the disk "...align=64" and XP
would not install. I see another post that says "...align=32" so here
I go to try that.

 

[Paul]

...


From the same forum... I tried aligning the disk "...align=64" and XP
would not install. I see another post that says "...align=32" so here
I go to try that.

Would it be easier to install on an ordinary hard drive, then prep
the SSD (set align=128) and then move the data over ?

http://www.ocztechnologyforum.com/forum/showthread.php?t=48309

Paul

 

[John Doe]

John Doe wrote:

[Cannot get Windows XP to install]

....

Would it be easier to install on an ordinary hard drive,

Many things would be easier.

http://www.ocztechnologyforum.com/forum/showthread.php?t=48309

I languished at the DOS prompt for about 24 hours,
formatting/aligning, and investigating. At least there were no signs
of disk corruption or failure, but Windows XP simply would not
install... The Windows XP installation CD persistently coughed up this
message "This disk does not contain a Windows XP-compatible
partition". In fact, the easy solution here was to switch the SATA
data cables. Now Windows XP installs just fine, and backup copies can
be restored to the SSD.

If aligning the OCZ Vertex partition really needs doing, using Windows
7 recovery utility Diskpart makes it easy.

[also taken from the OCC forum]

0 get to the command prompt

1 "diskpart"

2 "list disk"

3 "select disk X" (with X being the number given for your SSD)

4 "clean"

5 "create partition primary align 64"

6 "active"

7 "exit"

8 reboot, format that partition, and install XP

 

[John Doe]

Without tweaks...

copy 2.13 GB file from raptor to itself
61 seconds
from the raptor to the SSD
27 seconds
from the SDD to itself
87 seconds

 

[John Doe]

....

If aligning the OCZ Vertex partition really needs doing

I would think that particular tweak would show up in benchmarks, but
there was no difference whatsoever.

My Vertex SSD is labeled "1275". Maybe it was already updated to the
most recent version, published about three weeks ago. The drive felt
kind of warm... haha.

 

[Daniel Prince]

It has been more than five months since anyone has posted anything
in this thread. Has anyone learned anything new?

Do SSD's speed up Window XP significantly? Does XP write to one
area of the drive so much as to lower the drive life excessively? Do
any of these drives have a long (non-prorated) warranty? Has anyone
had a SSD fail? It so, how long did you use it and what did you use
it for?

How big of a SS drive is needed to speed up Windows XP Home
significantly? Are there any SSD's that should be avoided? Are
there any SSD's that should be sought out? Thank you in advance for
all replies.

 

[Ken]

Are there any SSD's that should be avoided?

Yes.
http://www.tomshardware.com/news/ssd-jmicron-jmf602,7057.html

 

[Guest]

Yes.
http://www.tomshardware.com/news/ssd-jmicron-jmf602,7057.html

The article mentions wear leveling, but doesn't explain that this
feature, in its current implementation, is far from perfect.

Windows is constantly reading and writing certain information
(registry files, pagefile etc), and because of this, SSDs need to
be replaced on a regular basis, certainly every year and possibly
more often if a system is heavily used.

 

[John Doe]

Has anyone learned anything new?

When I bought an OCZ SSD, the updated firmware was important for
good performance. Now it probably is current enough.

Do SSD's speed up Window XP significantly?

With updated firmware and tweaks, Maybe. Some things are
noticeably faster, for example working in Add or Remove Programs
is much faster for some reason, there is no longer a pause after a
program is removed.

Does XP write to one area of the drive so much as to lower the
drive life excessively?

Personally, I think that is a silly concern when a device has a
Mean Time between Failure (MTBF) of over one million hours.

Has anyone had a SSD fail?

Mine is doing just fine.

How big of a SS drive is needed to speed up Windows XP Home
significantly?

Same OS here. My SSD is only 31GB. I use it for the main hard
drive and make hidden copies of Windows to my Raptor. It works
fine. Unfortunately, apparently the price of SSD drives has not
decreased.

I am definitely not thrilled by the experience (it might help to
avoid stuttering for some gaming, but I have played little
lately). I do like the fact that the SSD uses much less power.
Some might like the fact that it is perfectly quiet.

If you have a serious business need, buy an SSD drive. You can
appreciate the fact that hard drive performance makes a little
difference that adds up during the day, or you can put the money
into something that clearly makes a difference. To buy or not to
buy depends on what you need it for versus your other options.

Of course, others have other opinions. Mine is from personal
experience with one SSD Drive since the date of my original post.

Good luck and have fun.

 

[John Doe]

The article mentions wear leveling, but doesn't explain that
this feature, in its current implementation, is far from
perfect.

And you, whoever you are, do not explain what "far from perfect"
is. Also, what you say there should follow your next point,
instead of the way you put it.

Windows is constantly reading and writing certain information
(registry files, pagefile etc), and because of this, SSDs need
to be replaced on a regular basis, certainly every year and
possibly more often if a system is heavily used.

Says who?

The SSD drive makers claim a greater than one million hours Mean
Time between Failure (MTBF). Obviously someone is lying.

















--

 

[Guest]

And you, whoever you are, do not explain what "far from perfect"
is. Also, what you say there should follow your next point,
instead of the way you put it.


Says who?

The SSD drive makers claim a greater than one million hours Mean
Time between Failure (MTBF). Obviously someone is lying.

Look a little closer at the specs, or simply read up on the subject:

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

The standard lifespan for individual segments on an SSD drive is
1000 write cycles. That's nothing compared to traditional hard
drives. The referenced article explains how SSD manufacturers
have tried to address this problem.

 

[John Doe]

Look a little closer at the specs,

What the **** are you talking about?

or simply read up on the subject:

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

The standard lifespan for individual segments on an SSD drive is
1000 write cycles.

And when recently did you learn about USENET?
In other words... Hopefully you are not a regular here.
Do not believe everything you read on a wikishit web site.
Nice to see someone who is eager to learn, but please just ask
instead of trolling for answers when you do not know something.

And please stop trying to change the follow-up groups after you
post.





















--

 

[Guest]

What the **** are you talking about?


And when recently did you learn about USENET?
In other words... Hopefully you are not a regular here.
Do not believe everything you read on a wikishit web site.
Nice to see someone who is eager to learn, but please just ask
instead of trolling for answers when you do not know something.

Concession noted. Thanks for playing.

 

[ohaya]

And you, whoever you are, do not explain what "far from perfect"
is. Also, what you say there should follow your next point,
instead of the way you put it.




Says who?

The SSD drive makers claim a greater than one million hours Mean
Time between Failure (MTBF). Obviously someone is lying.

Hi,

I think that you're confusing the device (SSD) reliability specification
(MTBF) with the device specifications for the number of writes.

Most (all, I think) SSDs use a bunch of some kind of non-volatile memory
chips. Those memory chips are such that they are limited as to the
number of times they can be written-to. That's why they have a
specification for the number of writes or write-cycles.

The MTBF spec is a kind of average (it's not exactly an average, but
kind of) of how long the overall device should work.

So, let's say you have a device (SSD) that has an MTBF of 1 million
hours, and a number of writes spec of 10000 write-cycles.

If you use that in your machine, and used it only for reading, the MTBF
spec is what should apply.

However, if you're writing to the device, then the number of writes spec
is what applies.

I hope that this helps clarify some things...

Jim

 

[Guest]

Hi,

I think that you're confusing the device (SSD) reliability specification (MTBF) with the device specifications for the number of
writes.

Most (all, I think) SSDs use a bunch of some kind of non-volatile memory chips. Those memory chips are such that they are limited
as to the number of times they can be written-to. That's why they have a specification for the number of writes or write-cycles.

The MTBF spec is a kind of average (it's not exactly an average, but kind of) of how long the overall device should work.

So, let's say you have a device (SSD) that has an MTBF of 1 million hours, and a number of writes spec of 10000 write-cycles.

If you use that in your machine, and used it only for reading, the MTBF spec is what should apply.

However, if you're writing to the device, then the number of writes spec is what applies.

I hope that this helps clarify some things...

Jim

That's exactly correct, Jim. MTBF is for the device, not for individual
R/W segments on the device. A SSD has no moving parts, so it's
MTBF spec is very misleading.

 

[John Doe]

I think that you're confusing the device (SSD) reliability
specification (MTBF) with the device specifications for the
number of writes.

Says who?

Most (all, I think) SSDs use a bunch of some kind of
non-volatile memory chips.

Was the door you came through to get to USENET labeled "CLUELESS
NEWBIES ENTER HERE"?

Those memory chips are such that they are limited as to the
number of times they can be written-to. That's why they have a
specification for the number of writes or write-cycles.

That is wrenchingly boring IMO.

The MTBF spec is a kind of average (it's not exactly an average,
but kind of) of how long the overall device should work.

Says another clueless newbie trolling for answers?

The MTBF specification might be wrong, but your justification for
the difference between the specification and the allegedly low
performance is nonsense.

Provide some authoritative citations.



















--

 

[Guest]

Says who?


Was the door you came through to get to USENET labeled "CLUELESS
NEWBIES ENTER HERE"?


That is wrenchingly boring IMO.


Says another clueless newbie trolling for answers?

The MTBF specification might be wrong, but your justification for
the difference between the specification and the allegedly low
performance is nonsense.

Provide some authoritative citations.

http://www.eettaiwan.com/STATIC/PDF/200808/EETOL_2008IIC_Spansion_AN_13.pdf

"If the system updates the 4 MB file 50 times per day then the system
(Flash) life expectancy would be ~1 day. This case shows the frequent
erasures of the same physical sector directly impacting the life
expectancy of the Flash device. Obviously a half day product life cycle
is not acceptable for typical applications."

Or in other words, it's only because of wear leveling that SSDs have
become usable at all.

 

[John Doe]

Added (sci.electronics.design), please feel free to trim groups.

"ohaya" <ohaya cox.net> wrote in message

That's exactly correct, Jim.

That is just amazing, Jack...

MTBF is for the device, not for individual R/W segments on the
device.

So you think the read/write "segments" on the device are different
than the device, Jack?

A SSD has no moving parts, so it's MTBF spec is very misleading.

BWAAAAAHAHAHAHAHAAAAAAA!!!

Please... Somebody, anybody, provide a citation that suggests
solid-state device MTBF figures are less accurate than physical
devices. I have a difficult time believing anyone could be that
much of a bullshit artist.

Thanks.

For any new readers... The argument is not that SSD device MTBF
figures are accurate. Currently, the argument is about some
apparently bizarre reasons why the MTBF figures are inaccurate.
The figures might be inaccurate, but more evident is that the two
quoted authors are clueless morons with zero credibility.



















--

 

[John Doe]

http://forums.storagereview.net/index.php?showtopic=27010

Be sure to read the referenced links

You want me to sift through opinions that could very easily have
been posted by A ****ING MORON LIKE YOU JACK, and you want me to
follow the links instead of you posting a relevant link here?

BWAAAAAHAHAHAHAHAAAAAAA!!!























--

 

[John Doe]

http://www.eettaiwan.com/STATIC/PDF/200808/EETOL_2008IIC_Spansion_AN_13.pdf

"If the system updates the 4 MB file 50 times per day then the
system (Flash) life expectancy would be ~1 day. This case shows
the frequent erasures of the same physical sector directly
impacting the life expectancy of the Flash device. Obviously a
half day product life cycle is not acceptable for typical
applications."

Or in other words, it's only because of wear leveling that SSDs
have become usable at all.

Yes of course flash memory has write limits, but you are taking a
major leap from that to your conclusion.

In other words... YOU ARE A ****ING MORON, JACK.



















--