Speed up my Windows XP Pro. SP3 with Flash USB drives/drives?

In ,
Twayne typed on Mon, 9 Nov 2009 11:31:23 -0500:
In ,
Ken Blake, MVP typed:
On Sun, 8 Nov 2009 12:59:40 -0700, "Bill in Co."
wrote:

Ken Blake, MVP wrote:
On Sun, 8 Nov 2009 00:23:40 -0500, "WMB"
wrote:

If speed is your objective, and cost is not a deterrent, image
your current
drive, install a SSD HD, re-install the image. "Chris Prillo" did
and says
its a slam bam for performance. I would sure like to hear from an
average
joe who tried it.


SSD drives tend to be on the small side and expensive. I have three
drives on my personal machine here. The C: drive is an
OCZSSD2-1VTX120G Vertex Series 120GB SATA II SSD ($370), and
is the place where Windows and applications are stored. The other
two drives are hard drives.

Yes, it seems very fast, but I don't have a previous experience on
this system to compare it with directly.

Ken, I had thought there was an issue with the fairly limited number
of max write cycles with any SSDs, making them NOT very suitable for
windows (like the regular C: drive, where there is always a LOT of
constant write activity).

(yes, I'm aware SSDs have some built-in "wear leveling" techniques
to spread this activity out over different arrays of storage
locations, but even at that...)


There may eventually be such an issue, but I haven't experienced it
so far. Perhaps I'm taking my chances with it, but I've been very
happy with the performance.

From what I've read it'll happen over a relatively short period of
time compared to mechanicals but you also get some extra time out of
it because it quits using the bad "sectors" and moves over to other
good ones. That goes on until there's no space left unless you're
watching it. I know of some SSD drives in a CT business (UTC) where
my son works, where they're being used but not on the system drives;
so far not a problem anywhere. They're surprisingly cheap bought in
quantity which tells us, I think, prime time isn't too far off.
They're using 64 Gig drives right now; really tiny in size!

Twayne`

Actually if you can believe the specs of SSDs, a manufacture claims the
MTBF is 227 years. And HDD are only less than 37 years.

Endurance:

SSD: MTBF 2 Million Hours
HDD: MTBF 300,000 Hours

http://talkback.zdnet.com/5208-11408...5468&s tart=0

I figured that I would have to overwrite a whole SSD, 24 times a day
(which would be hard to do without trying too). And it would take 11
years to wear one out.

--
Bill
Gateway MX6124 ('06 era) - Windows XP SP2

On Sun, 8 Nov 2009 20:26:05 -0600, "BillW50" wrote:

In ,
Ken Blake, MVP typed on Sun, 08 Nov 2009 18:55:10 -0700:
On Sun, 8 Nov 2009 17:03:05 -0600, "BillW50" wrote:

In ,
Ken Blake, MVP typed on Sun, 08 Nov 2009 13:11:50 -0700:
On Sun, 8 Nov 2009 12:59:40 -0700, "Bill in Co."
wrote:
Ken, I had thought there was an issue with the fairly limited
number of max write cycles with any SSDs, making them NOT very
suitable for windows (like the regular C: drive, where there is
always a LOT of constant write activity).

(yes, I'm aware SSDs have some built-in "wear leveling" techniques
to spread this activity out over different arrays of storage
locations, but even at that...)

There may eventually be such an issue, but I haven't experienced it
so far. Perhaps I'm taking my chances with it, but I've been very
happy with the performance.

Nor all SSD are equal for one. For example:

1) SLC type lasts 100,000 or more complete writes

2) MLC type lasts 5,000 to 10,000 complete writes

Secondly if you want to limit the number of writes, this is easy to
do. Just turning off System Restore and a swapfile is a very big
help right there.

Thirdly, you can totally stop all writes if you want to by merging MS
EWF files into your Windows XP.

Thanks for the info. Do you know whether my OCZSSD2-1VTX120G Vertex
Series 120GB SATA II SSD is SLC or MLC?

A Google search seems to suggest it is a MLC type. But I don't know for
sure. At first, SLC type was used for SSDs, but they were really
expensive. Since then MLC type has been improving in technology and half
the price to manufacture. Plus SLC types are disappearing from the
marketplace recently.

I have seen heavy use of MLC SSD die in a year or two. They usually
have a warrantee of one year. SLC SSD should last 10 years plus
without problems. Although if you limit writing, you will be long
dead before you ever wear one of them out.

Assuming that mine is SLC, and noting that I'm 72 years old, I *may*
be long dead before I wear it out.

Yes and I am about 20 years behind you. And I think the computer with
SSD will most likely be so outdated that you wouldn't use it anyway
before it fails.


I typically get 4-5 years usage out of a computer before I replace it.
This one is brand new, so I'll probably be 76-77 when I want to
replace it.



Unless it is a really cheap MLC type. Although yours
seems to be one of the better ones. ;-)

--
Bill
Gateway MX6124 ('06 era) - Windows XP SP2


--
Ken Blake, Microsoft MVP (Windows Desktop Experience) since 2003
Please Reply to the Newsgroup

On Mon, 9 Nov 2009 11:22:34 -0600, "BillW50" wrote:


Actually if you can believe the specs of SSDs, a manufacture claims the
MTBF is 227 years. And HDD are only less than 37 years.


I'm always reluctant to believe MTBF claims. They may be correct, but
with numbers as high as 227 years, I have no confidence in them.

--
Ken Blake, Microsoft MVP (Windows Desktop Experience) since 2003
Please Reply to the Newsgroup

In ,
Ken Blake, MVP typed on Mon, 09 Nov 2009 11:28:51 -0700:
On Mon, 9 Nov 2009 11:22:34 -0600, "BillW50" wrote:


Actually if you can believe the specs of SSDs, a manufacture claims
the MTBF is 227 years. And HDD are only less than 37 years.


I'm always reluctant to believe MTBF claims. They may be correct, but
with numbers as high as 227 years, I have no confidence in them.

Same here, but I have done the math. If every cell of a SSD can be
written to 100,000 times, it would take a person overwriting the whole
SSD 24 times a day for 11 years before you would wear one out. That is a
lot of writing. So I can see the average user might get 227 years out of
one.

--
Bill
Gateway MX6124 ('06 era) - Windows XP SP2

BillW50 wrote:
In ,
Ken Blake, MVP typed on Mon, 09 Nov 2009 11:28:51 -0700:
On Mon, 9 Nov 2009 11:22:34 -0600, "BillW50" wrote:


Actually if you can believe the specs of SSDs, a manufacture claims
the MTBF is 227 years. And HDD are only less than 37 years.


I'm always reluctant to believe MTBF claims. They may be correct, but
with numbers as high as 227 years, I have no confidence in them.

Same here, but I have done the math. If every cell of a SSD can be
written to 100,000 times, it would take a person overwriting the whole
SSD 24 times a day for 11 years before you would wear one out. That is a
lot of writing. So I can see the average user might get 227 years out of
one.

So I guess the bottom line is it sounds like there's really no issue with
using the SSDs to replace conventional HDs except for the price (no matter
which type, but skipping the DRAM ones, which don't seem useful for the
general consumer).

IOW, they will outlast any conventional HD (no matter what type of SSD), and
are certainly a lot faster. I'm still not sure about the permanence of of
the data stored in flash memory in terms of its shelf life (or maybe that
was expressed in its MTBF stats), but I'm guessing that's not an real issue,
either, in comparison to the mechanical drives.

In ,
Bill in Co. typed on Mon, 9 Nov 2009 13:23:07 -0700:
BillW50 wrote:
In ,
Ken Blake, MVP typed on Mon, 09 Nov 2009 11:28:51 -0700:
On Mon, 9 Nov 2009 11:22:34 -0600, "BillW50"
wrote:
Actually if you can believe the specs of SSDs, a manufacture claims
the MTBF is 227 years. And HDD are only less than 37 years.


I'm always reluctant to believe MTBF claims. They may be correct,
but with numbers as high as 227 years, I have no confidence in them.

Same here, but I have done the math. If every cell of a SSD can be
written to 100,000 times, it would take a person overwriting the
whole SSD 24 times a day for 11 years before you would wear one out.
That is a lot of writing. So I can see the average user might get
227 years out of one.

So I guess the bottom line is it sounds like there's really no issue
with using the SSDs to replace conventional HDs except for the price
(no matter which type, but skipping the DRAM ones, which don't seem
useful for the general consumer).

IOW, they will outlast any conventional HD (no matter what type of
SSD), and are certainly a lot faster. I'm still not sure about the
permanence of of the data stored in flash memory in terms of its
shelf life (or maybe that was expressed in its MTBF stats), but I'm
guessing that's not an real issue, either, in comparison to the
mechanical drives.

Yes that is about it. Except I wouldn't call the real cheap MLC SSD (vs.
more expensive MLC SSD and better yet SLC SSD), as fast. As they do make
some really slow MLC types. For example Super Talent has one called
FPM16GHAE PATA PCIe SSD with a read speed of 45M and a write of 15MB.

My personal experience is that HDD has an early failure rate of 3 out of
21. And they lasted a month or less. SSDs are hitting the same ratio, 1
out of 7. And they usually fail in a few months. And this one will work
if you let it sit with power for 90 minutes. So remember this trick if
you ever have one fail. And there might have been a recall on that lot.
And I believe it was a failed controller on the SSD board and not any of
the SSD chips themselves.

--
Bill
Gateway MX6124 ('06 era) - Windows XP SP2

In ,
Ken Blake, MVP typed:
On Mon, 9 Nov 2009 11:22:34 -0600, "BillW50" wrote:


Actually if you can believe the specs of SSDs, a manufacture claims
the MTBF is 227 years. And HDD are only less than 37 years.


I'm always reluctant to believe MTBF claims. They may be correct, but
with numbers as high as 227 years, I have no confidence in them.

MTBFs are a little like statistics; you can make them say different things
at different times. Folks should remember the calcs for MTBFs are usually
for laboratory conditions of a controlled environment, etc. etc.. I
wouldn't expect to see an SSD drive last 200 years in use any more than I
would a HDD last 37 years in use. Whether it's alpha migration or physical
parts wear, neither comes out very "real" in MTBF calcs. Grease dries,
irradiation degrades and all that good stuff. I've never seen a single
instance of a product making its MTBF numbers, have you? MTBF is really
only useful (sometimes) as a comparison factor, not real numbers.
Like in the G's tests, you have no idea how the products were dropped if
they don't say so and there are many different ways of speccing how to
calculate (not measure) them.
I used to have to do a lot of MTBF on my designs and I hated it; it felt
like lying to the customers because marketing always hyped it as how long
the product would last, which it definitely is not. You CAN do actual-use
calcs for MTBFs, but it's expensive and time consuming so all most engineers
do is use the means of the various parts within a product. It has its uses,
but not in advertising, marketing or anything for the public, in reality.

Cheers,

Twayne`

Actually, HD MTBF numbers are PURE statistics!

Twayne wrote:

In ,
Ken Blake, MVP typed:

On Mon, 9 Nov 2009 11:22:34 -0600, "BillW50" wrote:



Actually if you can believe the specs of SSDs, a manufacture claims
the MTBF is 227 years. And HDD are only less than 37 years.


I'm always reluctant to believe MTBF claims. They may be correct, but
with numbers as high as 227 years, I have no confidence in them.


MTBFs are a little like statistics; you can make them say different things
at different times. Folks should remember the calcs for MTBFs are usually
for laboratory conditions of a controlled environment, etc. etc.. I
wouldn't expect to see an SSD drive last 200 years in use any more than I
would a HDD last 37 years in use. Whether it's alpha migration or physical
parts wear, neither comes out very "real" in MTBF calcs. Grease dries,
irradiation degrades and all that good stuff. I've never seen a single
instance of a product making its MTBF numbers, have you? MTBF is really
only useful (sometimes) as a comparison factor, not real numbers.
Like in the G's tests, you have no idea how the products were dropped if
they don't say so and there are many different ways of speccing how to
calculate (not measure) them.
I used to have to do a lot of MTBF on my designs and I hated it; it felt
like lying to the customers because marketing always hyped it as how long
the product would last, which it definitely is not. You CAN do actual-use
calcs for MTBFs, but it's expensive and time consuming so all most engineers
do is use the means of the various parts within a product. It has its uses,
but not in advertising, marketing or anything for the public, in reality.

Cheers,

Twayne`

Bob I wrote on Tue, 10 Nov 2009 07:45:33 -0600:
Actually, HD MTBF numbers are PURE statistics!

Actually if less than 37 years holds true, I have some HDD that still
have half of their life to go. And they are still working just fine the
last time I checked. ;-)

--
Bill
Asus EEE PC 702G4 ~ 2GB RAM ~ 16GB-SDHC
Xandros Linux (build 2007-10-19 13:03)

Twayne wrote on Mon, 9 Nov 2009 11:31:23 -0500:
From what I've read it'll happen over a relatively short period of time
compared to mechanicals but you also get some extra time out of it because
it quits using the bad "sectors" and moves over to other good ones. That
goes on until there's no space left unless you're watching it. I know of
some SSD drives in a CT business (UTC) where my son works, where they're
being used but not on the system drives; so far not a problem anywhere.
They're surprisingly cheap bought in quantity which tells us, I think, prime
time isn't too far off. They're using 64 Gig drives right now; really tiny
in size!

Twayne`

Actually all SSD that I know of uses wear leveling. This increases the
number of writes, but the whole mass storage is written to evenly. Thus
at the end of its life, the whole thing just dies. Too bad they don't
add a counter or something to let you know how far you are in its life
cycle. ;-)

--
Bill
Asus EEE PC 702G4 ~ 2GB RAM ~ 16GB-SDHC
Xandros Linux (build 2007-10-19 13:03)

In ,
Bob I typed:
Actually, HD MTBF numbers are PURE statistics!

I was trying to remember that very name! lol, thanks!
Twayne`


Twayne wrote:

In ,
Ken Blake, MVP typed:

On Mon, 9 Nov 2009 11:22:34 -0600, "BillW50"
wrote:
Actually if you can believe the specs of SSDs, a manufacture claims
the MTBF is 227 years. And HDD are only less than 37 years.


I'm always reluctant to believe MTBF claims. They may be correct,
but with numbers as high as 227 years, I have no confidence in them.


MTBFs are a little like statistics; you can make them say different
things at different times. Folks should remember the calcs for
MTBFs are usually for laboratory conditions of a controlled
environment, etc. etc.. I wouldn't expect to see an SSD drive last
200 years in use any more than I would a HDD last 37 years in use.
Whether it's alpha migration or physical parts wear, neither comes
out very "real" in MTBF calcs. Grease dries, irradiation degrades
and all that good stuff. I've never seen a single instance of a
product making its MTBF numbers, have you? MTBF is really only
useful (sometimes) as a comparison factor, not real numbers. Like
in the G's tests, you have no idea how the products were dropped if
they don't say so and there are many different ways of speccing how
to calculate (not measure) them. I used to have to do a lot of
MTBF on my designs and I hated it; it felt like lying to the
customers because marketing always hyped it as how long the product
would last, which it definitely is not. You CAN do actual-use calcs
for MTBFs, but it's expensive and time consuming so all most
engineers do is use the means of the various parts within a product.
It has its uses, but not in advertising, marketing or anything for
the public, in reality. Cheers,

Twayne`

In message , BillW50
writes:
[]
Actually all SSD that I know of uses wear leveling. This increases the
number of writes, but the whole mass storage is written to evenly. Thus
at the end of its life, the whole thing just dies. Too bad they don't
add a counter or something to let you know how far you are in its life
cycle. ;-)

Do they not have the equivalent of SMART?
--
J. P. Gilliver. UMRA: 1960/1985 MB++G.5AL-IS-P--Ch++(p)Ar@T0H+Sh0!:`)DNAf
** http://www.soft255.demon.co.uk/G6JPG-PC/JPGminPC.htm for ludicrously
outdated thoughts on PCs. **

Britain is still a class-ridden society. As soon as a man opens his mouth, we
can tell in what sort of school he missed his education. (George Mikes, "How to
be Decadent" [1977].)

In ,
J. P. Gilliver (John) typed on Fri, 13 Nov 2009 07:27:08 +0000:
In message , BillW50
writes:

Actually all SSD that I know of uses wear leveling. This increases
the number of writes, but the whole mass storage is written to
evenly. Thus at the end of its life, the whole thing just dies. Too
bad they don't add a counter or something to let you know how far
you are in its life cycle. ;-)

Do they not have the equivalent of SMART?

Actually they do use SMART. Although SMART doesn't have SSD useful
information as a standard yet. Programs like Hard Disk Sentinel does
help in this regard though. As it keeps a running tally on how much is
written to the SSD.

--
Bill
Asus EEE PC 702G8 ~ 2GB RAM ~ 16GB-SDHC
Windows XP SP2

I'd be very interested in going SSD as my main drive, but the lifespan is
always what scares me. I want my data to be relatively safe. Being that the
technology is newer than a hard drive is what worries me since I have no
experience with it.

SMART attributes aren't always a good measure of a drive I'm told. Different
programs read differently. I'm monitoring my drives at work and they list
33% health left. I still had a Windows 98 machine running here up until a
few months ago with the original hard drive.

"WMB" wrote in message
...
If speed is your objective, and cost is not a deterrent, image your
current drive, install a SSD HD, re-install the image. "Chris Prillo" did
and says its a slam bam for performance. I would sure like to hear from
an average joe who tried it.

"Ant" wrote in message
...
Um, nice poem?


On 8/16/2009 2:58 PM PT, db typed:

there is a third party
maker that makes
a ready boost version
for xp.

it's about 50 bucks.

-------------

they really act like
the hibernation feature

and you require usb
flash drives that are
twice the size of your
ram to provide the
full benefit.

also, flash drives come
in two flavors:

those that are ready
boost ready

and those that are
not.

-------------

one day microsoft
will hire the smart guy
who develops a rom
level hibernation.

but it will be a long
time til then.
--
"Though your enemy is the size of an ant, look upon him as an
elephant." --Danish
/\___/\
/ /\ /\ \ Phil/Ant @ http://antfarm.ma.cx (Personal Web Site)
| |o o| | Ant's Quality Foraged Links (AQFL): http://aqfl.net
\ _ / Nuke ANT from e-mail address: NT
( ) or
Ant is currently not listening to any songs on his home computer.

In ,
shawn typed on Wed, 30 Dec 2009 14:34:14 -0500:
I'd be very interested in going SSD as my main drive, but the
lifespan is always what scares me. I want my data to be relatively
safe. Being that the technology is newer than a hard drive is what
worries me since I have no experience with it.

Hi Shawn! MTBF for SSD is 227 years. While hard drives MTBF is 37 years.

SMART attributes aren't always a good measure of a drive I'm told.
Different programs read differently. I'm monitoring my drives at work
and they list 33% health left. I still had a Windows 98 machine
running here up until a few months ago with the original hard drive.

Google research showed that SMART is unreliable. I take it as a small
gauge. But not something you should depend on 100%. As 40% of failures
are not detected by SMART.

--
Bill
Gateway MX6124 ('06 era) 1 of 3 - Windows XP SP2