The HP company sells computers with motherboards and BIOS
set the way they want. HP does not want to deal with
problems caused by tinkering, so they limit your tinkering
options.
Just to point out how far they went in the past (your
conditions may vary), the first Windows computer I bought
was a HP 6465 with a 433 Celeron. Later as I learned more
about computers I found that HP used a ASUS MEV-BM
motherboard with an Intel 440BX chipset. The ASUS mobo
manual shows many options in setting the mobo jumpers to set
the FSB and CPU multiplier. Back in those days Intel wanted
all the Celerons to run only on a 66 MHz bus. But since the
MEV-BM mobo and all the other components were designed to
run faster I wanted to set the jumpers to a 100 MHz FSB and
the CPU multiplier at 4.5 or 5 so the CPU would not be
over-clocked. But HP had used flush cutting tools to remove
all the jumpers except the ones used to set the FSB to 66
MHz and the CPU at 6.5.
With the new mobo using the BIOS to control everything, you
might be able to get a standard BIOS for the motherboard,
but you recovery CDs won't work without the oem BIOS.
If you want to be able to tweak, build your own and pick
your components (or have one custom assembled if you're
afraid of small tools and simple directions).
--
The people think the Constitution protects their rights;
But government sees it as an obstacle to be overcome.
| Your answer does indeed seem very knowledgeable and nearly
resolves my
| question. Many thanks.
|
| It would seem, however, that the key to wringing extra
performance from the
| DDR333 memory would be in whether the BIOS supplied with
the zt3000/735
| actually allowed the memory to be underclocked, or even
configured at all.
| Perhaps someone at HP would know. As for myself, I barely
understand the
| technique.
|
| So, without a look at that BIOS, it seems that wringing
any extra
| performance out of the PC2700 would be a long-shot. And,
given the price
| differential, perhaps not cost-effective.
|
| I am tempted to believe that the best increment in power
is just to stuff
| the computer with the full 2x1GB of memory, even at 288
MHz. I am tempted,
| but a tad reluctant, since I believe that each of these
1GB SODIMMs draws
| down a full watt to keep itself refreshed (whacking your
battery performance
| sharply); and also the system then requires
correspondingly large standby
| recovery images and swap files. Perhaps that
configuration is overdoing it
| also.
|
| Any thoughts?
|
| Thanks for the excellent information!
|
|
| | > Here is the definitive answer for you. The new boards
with support for
| > the Dothan core are generally supporting PC2700. The
major performance
| > increase is really going to be in the Dothan core itself
as you have a
| > much larger on-die cache and a much better pipe-staging
for the Dothan
| > core. As for the PC2700 vs. PC2100, it is true that if
you are limited
| > to 266 MHz, then 333 MHz memory will only run at that
speed, but there
| > is a catch. You may be able to get the 333 MHz memory
to run faster
| > because you can lower the timings on the memory. A
timing of 2-2-2-6
| > would be possible by a 333 MHz memory chip underclocked
to 266, whereas,
| > a 266 may not be able to reach those latency timings.
| >
| > The downside here is that most OEM computers (Dell, HP,
etc.) have
| > limited BIOS that will not allow you to change these
settings. I have a
| > Dell Latitude D800, and it has the Pentium M 735 with
DDR333 memory in
| > it. It runs great, and I have not had any problems
except no possibility
| > to overclock
| >
| > Nathan McNulty
| >
| > Nick Burns wrote:
| >
| > > As long as you are running at 266, you are not going
to see improvement.
| > >
| > > | > >
| > >>A recent thread in this group (started by Rodney
Chelius) discussed
| > >
| > > whether
| > >
| > >>it was necessary or advantageous to use PC2700 memory
(333 MHz) vs.
| PC2100
| > >>memory (266 MHz) on a HP Pavilion zt3000, which has a
Pentium-M CPU.
| The
| > >>conclusion seemed to be that the motherboard limited
the speed to 266,
| and
| > >>so there was no advantage to the faster memory.
| > >>
| > >>Since that discussion was some months ago, and it is
now possible to get
| a
| > >>2.0 GHz Dothan Pentium-M on this model, I wonder if
this still holds
| true.
| > >>That is, are the faster models still built on the same
motherboard and
| > >
| > > still
| > >
| > >>limited to 266 MHz? Perhaps it is a dumb question,
but I wonder if the
| > >>Banias mobo is different from the Dothan?
| > >>
| > >>TNX
| > >>
| > >>
| > >
| > >
| > >
|
|
