XP Home on P-III 700

[William Kossack]

I would like to replace windows 98 with XP or XP home in the house.

I'm debating rebuilding my daughters socket 370 celeron 600 giving her
my old P-III 700.

Would there be an appreciable performance improvment with the 700?

I think I may have an extra stick of RAM that might go onto her Mboard
so I could up her RAM to 512MB.

 

[Charlie]

I run a PIII 700 with 512 RAM and it runs great on XP Home. I made it even
better by buying a cheap 40 gig 7200 RPM disk drive from Newegg and have
been completely satisfied. I can't say for sure but I believe your
performance improvement will be worth doing.

 

[Guest]

I run a PIII 700 with 512 RAM and it runs great on XP Home. I made it even
better by buying a cheap 40 gig 7200 RPM disk drive from Newegg and have
been completely satisfied. I can't say for sure but I believe your
performance improvement will be worth doing.

Go for it if you want the experience. The performance improvement will be
minimal. My second computer is close to what your are porposing and I agonize
when using it. However, for just a few more $:

Essential Technology on a Budget
Bullet Intel® Pentium® 4 Processor (2.80GHz, 533 FSB)
Bullet Microsoft® Windows® XP Home Edition
Bullet 256MB DDR SDRAM at 400MHz
Bullet 40GB Ultra ATA/100 7200RPM Hard Drive
Bullet FREE 2-Day Shipping - Online Only!
Bullet FREE 2nd Bay CD-Burner - Online Only!

Dimension 4700
Enhanced Performance
Bullet Pentium® 4 Processor 530 with HT Technology (3.00GHz, 800 FSB)
Bullet Microsoft® Windows® XP Home Edition
Bullet 256MB DDR2 SDRAM at 400MHz (1x256M)
Bullet 40GB Serial ATA Hard Drive (7200RPM)
Bullet FREE 2-Day Shipping - Online Only!
Bullet FREE 2nd Bay CD-Burner - Online Only!
Featured at
$549
$499
After $50 OFF Instantly!

 

[Ken Blake]

In

I would like to replace windows 98 with XP or XP home in the
house.

I'm debating rebuilding my daughters socket 370 celeron 600
giving her
my old P-III 700.

Would there be an appreciable performance improvment with the
700?

Are you asking where the P3-700 will provide an appreciable
performance improvement over a celeron 600?

No, I doubt it.

I think I may have an extra stick of RAM that might go onto her
Mboard
so I could up her RAM to 512MB.

From what? And what will apps she be running? Depending on the
answers to those questions, the extra RAM may not help
performance either.

 

[William Kossack]

This is a computer used by a preteenaged girl so there is mostly chat
and browsing and email and little gaming or anything that I would
consider to be a processor hog.

The current computer has 256MB of Ram but I could up it to 512mb.

I have enought working archaic stuff sitting around not being used that
has no value. I want to delay spending more money (XP Home will cost
enough).

 

[Winux P]

Hi William,

I have a PIII 550 with 512 MB RAM, WinXP Pro SP2 runs very well on it. Also
have an ATI x2 graphics card with 128 MB on the card. The only trade off is
I can't install\play nearly all modern games on it. Apart from that, PIII
700 can only be better.

Regards,
Winux P.

 

[Guest]

This is a computer used by a preteenaged girl so there is mostly chat
and browsing and email and little gaming or anything that I would
consider to be a processor hog.

The current computer has 256MB of Ram but I could up it to 512mb.

I have enought working archaic stuff sitting around not being used that
has no value. I want to delay spending more money (XP Home will cost
enough).

Go for the upgrade.

 

[Ken Blake]

In

This is a computer used by a preteenaged girl so there is
mostly chat
and browsing and email and little gaming or anything that I
would
consider to be a processor hog.

The current computer has 256MB of Ram but I could up it to
512mb.

Given what it will be used for, 256MB is probably sufficient.

More RAM will help performance only if it keeps her from using
the page file. I would delay any decision on installing more RAM,
and install XP. Then download and run Bill James's pagefile usage
monitor (you can find it at
http://www.kellys-korner-xp.com/xp_p.htm). If that shows
significant page file usage, then it makes sense to install
additional RAM; otherwise it doesn't.

 

[Ron Martell]

I would like to replace windows 98 with XP or XP home in the house.

I'm debating rebuilding my daughters socket 370 celeron 600 giving her
my old P-III 700.

Would there be an appreciable performance improvment with the 700?

I think I may have an extra stick of RAM that might go onto her Mboard
so I could up her RAM to 512MB.

Going to 512 mb of RAM will probably do more for the performance than
will the faster CPU, especially if the present configuration is only
256 mb of RAM or less.

Good luck


Ron Martell Duncan B.C. Canada
--
Microsoft MVP
On-Line Help Computer Service
http://onlinehelp.bc.ca

"The reason computer chips are so small is computers don't eat much."

 

[Alex Nichol]

I would like to replace windows 98 with XP or XP home in the house.

I'm debating rebuilding my daughters socket 370 celeron 600 giving her
my old P-III 700.

Would there be an appreciable performance improvment with the 700?

Probably not enough to notice. The extra RAM would probably do more,
provided it is an exact match (Windows is very fussy on this)

 

[Jo]

The extra RAM would probably do more,
provided it is an exact match (Windows is very fussy on this)

This is news to me. Please would you explain what you mean by 'exact
match' and why Windows is so fussy about it?

 

[Ron Martell]

This is news to me. Please would you explain what you mean by 'exact
match' and why Windows is so fussy about it?

RAM from different manufacturers, or even from the same manufacturer
but different production lots, can have internal differences in the
material composition and/or the physical layout of the RAM chips.
These differences can affect the strength and timing of the signals
returned from the RAM and Windows is sensitive to these differences.

The classic symptom of this sort of problem is where the computer
works flawlessly when either the old or the new RAM modules are
installed by themselves but there are problems and errors when both
old and new modules are installed together.

There is no fix or work-around, other than replacing one set of RAM
modules so that all of the installed RAM is identical.

Hope this clarifies the situation.

Good luck


Ron Martell Duncan B.C. Canada
--
Microsoft MVP
On-Line Help Computer Service
http://onlinehelp.bc.ca

"The reason computer chips are so small is computers don't eat much."

 

[Jo]

RAM from different manufacturers, or even from the same
manufacturer but different production lots...

That would suggest this is a *very* common problem yet I've never
even heard of it, let alone experienced it. Can you provide any
references to substantiate your assertion?

 

[Alex Nichol]

This is news to me. Please would you explain what you mean by 'exact
match' and why Windows is so fussy about it?

Windows includes some very tightly calibrated timing loops. These are
checked at boot - if they subsequently end up with a RAM module of even
slightly different characteristics they may crash.

 

[Alex Nichol]

That would suggest this is a *very* common problem yet I've never
even heard of it, let alone experienced it. Can you provide any
references to substantiate your assertion?

Not very many people add RAM modules of different characteristics, but
this is a problem *we* see regularly in these groups where someone has
added RAM and then says the system keeps crashing. The authority for
the explanation was Raymond Chen of the Windows developers

 

[Guest]

I can remember seeing this problem under WinNT on older computers (which had
earlier worked fine under Win31 or Win9x), but have rarely seen it on most
newer machines. I suspect that most memory now meets or exceeds any
specifications that matter, and so you can mix most of them without much
trouble. I even see mixing different speeds of memory (like PC100/133
or PC2100/2700/3200 - usually they all then run at the speed of the slowest
chips) without much problem. I have heard rumors of cases where it makes a
difference, but after servicing many hundreds of computers, I have rarely seen
any problems from mixing memory (unless you are doing silly things like
overclocking).

I have seen a few problems where it seems to make a difference which slot on
the motherboard you put the memory into. Actually, what seems to be the
factor is which slots you leave _empty_. Sometimes by moving the memory to
other slots, you can make problems go away. I'm not sure why this would be
the case. Perhaps just taking the memory out and in again results in cleaning
off the contacts in the socket (which is often something that helps -
epecially the AGP socket which has such small contacts).

|
|>|>
|>> The extra RAM would probably do more,
|>> provided it is an exact match (Windows is very fussy on this)
|>
|>This is news to me. Please would you explain what you mean by 'exact
|>match' and why Windows is so fussy about it?
|
|RAM from different manufacturers, or even from the same manufacturer
|but different production lots, can have internal differences in the
|material composition and/or the physical layout of the RAM chips.
|These differences can affect the strength and timing of the signals
|returned from the RAM and Windows is sensitive to these differences.
|
|The classic symptom of this sort of problem is where the computer
|works flawlessly when either the old or the new RAM modules are
|installed by themselves but there are problems and errors when both
|old and new modules are installed together.
|
|There is no fix or work-around, other than replacing one set of RAM
|modules so that all of the installed RAM is identical.
|
|Hope this clarifies the situation.
|
|Good luck
|
|
|Ron Martell Duncan B.C. Canada

 

[Jo]

Windows includes some very tightly calibrated timing loops

Yeah right. Windows copes fine with processor clock speed variation
of 40:1, different processors and chipsets, different amounts of L1
and L2 cache etc, but screws up on manufacturing tolerances of main
memory because of timing loops?

Sounds pretty unlikely to me.

 

[Ron Martell]

Yeah right. Windows copes fine with processor clock speed variation
of 40:1, different processors and chipsets, different amounts of L1
and L2 cache etc, but screws up on manufacturing tolerances of main
memory because of timing loops?

Sounds pretty unlikely to me.

It screws up because some parts of the information read from RAM
arrives a few nanoseconds before or after the rest of the information,
and/or because one part of the RAM signal rises to peak amplititude a
fraction of a nanosecond sooner than the other part.


Ron Martell Duncan B.C. Canada
--
Microsoft MVP
On-Line Help Computer Service
http://onlinehelp.bc.ca

"The reason computer chips are so small is computers don't eat much."

 

[Jo]

It screws up because some parts of the information read from RAM
arrives a few nanoseconds before or after the rest of the
information, and/or because one part of the RAM signal rises to
peak amplititude a fraction of a nanosecond sooner than the
other part.

I understand data skew problems on a parallel bus. A faulty or
overclocked RAM stick might well produce such a problem.

But
1) Why would mismatch *between* ram sticks (that are not faulty and
are both operating within spec) cause such a problem?

2) What has timing skew to do with a "tightly calibrated timing
loops" in Windows? [and does anyone still write such fragile code
these days anyway?]

 

[Ron Martell]

But
1) Why would mismatch *between* ram sticks (that are not faulty and
are both operating within spec) cause such a problem?

Because the different RAM can be made of different materials which
have different electrical characteristics and/or because differences
in the physical layout of the RAM modules can affect the length of the
electrical path and therefore the time that it takes the signal to
travel that path.

That is right. Computer speeds are now such that the time it takes an
electrical signal to travel a centimeter is a significant portion of a
CPU cycle on a multi-ghz processor.


Ron Martell Duncan B.C. Canada
--
Microsoft MVP
On-Line Help Computer Service
http://onlinehelp.bc.ca

"The reason computer chips are so small is computers don't eat much."