P4 MOBO

  • Thread starter Thread starter Claus Christensen
  • Start date Start date
C

Claus Christensen

Hi all.
A friend of mine would like to built his own new machine. Mostly for gaming.
AMD or Pentium ?? He would like a P4 about 2.0. What mobo should he get?
Must be 8*AGP .I built my own AMD with a FIC13 mobo, but i dont know all
that much about Pentium. And prob about 2gig of at least PC2100 memory.
All help would be appreciated.
Have a good Thanksgiving.
C
 
Claus said:
Hi all.
A friend of mine would like to built his own new machine. Mostly for
gaming. AMD or Pentium ?? He would like a P4 about 2.0. What mobo should
he get?

2.4C P4/800, asus P4P800 is a good start for a P4 system.
Must be 8*AGP .I built my own AMD with a FIC13 mobo, but i dont
know all
that much about Pentium. And prob about 2gig of at least PC2100 memory.

Why 2 gig of ram? You want DDR400 ram for that board and use 2 matching
sticks to use the dual chanel part of it.
 
Claus Christensen said:
Hi all.
A friend of mine would like to built his own new machine. Mostly for
gaming.
AMD or Pentium ??

Based on my recent personal experience watching an AMD600 on a
Gigabyte board and an AMD Thunderbird 800 on an Abit board ... and
both damned systems go bellly up on me and having never been terribly
stable, I've officially regretted the two AMD purchases I've made.
All this despite my having been very careful in building the 800
system, using AMD approved power supplies, branded Corsair memory, not
overclocking, and having a dual fan case that was as loud as a frakin
jet engine.

In addition, many folks are concerned about how hot the AMD systems
tend to run..which makes reliability and stability tougher nuts to
crack.
He would like a P4 about 2.0. What mobo should he get?

For what it's worth, I've had fabulous luck with my Asus
P4GE-V/LAN-UAY motherboard. It includes integrated video (which he
won't definitely won't want to use) as well as an AGP slot. There is
a version available without integrated video, I believe. Dunno about
8x. Every Asus built system I've had has been solid.

The other mobo manufacturer I'd look at is Intel..for obvious
reasons.
Must be 8*AGP .I built my own AMD with a FIC13 mobo, but i dont know
all that much about Pentium. And prob about 2gig of at least PC2100
memory. All help would be appreciated. Have a good Thanksgiving.
C

Best Regards,
 
Based on my recent personal experience watching an AMD600 on a
Gigabyte board and an AMD Thunderbird 800 on an Abit board ... and
both damned systems go bellly up on me and having never been terribly
stable, I've officially regretted the two AMD purchases I've made.

But that's 4 generations ago... MORE recently someone buying an Intel
P3 1.13GHz would have even more reason to avoid Intel. Point being,
the past is the past, problems get fixed.

All this despite my having been very careful in building the 800
system, using AMD approved power supplies, branded Corsair memory, not
overclocking, and having a dual fan case that was as loud as a frakin
jet engine.

... and if you look back at the problem reports, none of that mattered
because it was the motherboards more often than not. That problem has
been resolved.

In addition, many folks are concerned about how hot the AMD systems
tend to run..which makes reliability and stability tougher nuts to
crack.

Not true. Intel systems run hotter.
Repeat after me...
Modern and next-gen. Intel systems run hotter.
They "might" idle cooler, but not at full load. If you'll never run
at full load there is no reason for a modern system. It is a fact,
check the respective CPU datasheets.

For what it's worth, I've had fabulous luck with my Asus
P4GE-V/LAN-UAY motherboard. It includes integrated video (which he
won't definitely won't want to use) as well as an AGP slot. There is
a version available without integrated video, I believe. Dunno about
8x. Every Asus built system I've had has been solid.

The other mobo manufacturer I'd look at is Intel..for obvious
reasons.

This I agree with, Asus and Intel are two of the best.


Soounds a bit like a mismatched system... 2GB of PC2100 sounds right
for a server, but 512MB-1GB of PC3200 sounds more appropriate for a
P4-based "PC".



Dave
 
Todd said:
Based on my recent personal experience watching an AMD600 on a
Gigabyte board and an AMD Thunderbird 800 on an Abit board ... and
both damned systems go bellly up on me and having never been terribly
stable, I've officially regretted the two AMD purchases I've made.

$20 says BOTH were built with a board using a Via chipset.

It's not the processor, it's those damned Viachipsets that have turned many
people against AMD. I learned that Via chipsets systems can suck no matter
which CPU is used!
 
kony said:
But that's 4 generations ago... MORE recently someone buying an Intel
P3 1.13GHz would have even more reason to avoid Intel. Point being,
the past is the past, problems get fixed.

Shame is Via is STILL producing crap hardware years later..
 
stacey said:
$20 says BOTH were built with a board using a Via chipset.

You'd lose that bet I'm afraid.

The 600 was on a Ali Aladdin V AGP set on a Gigabyte GA-5AX.

The 800 was a Via KT133 chipset, which at the time was apparently
better refered than the Ali alternative. Abit KT7-RAID motherboard
(but not using RAID).

It's a lot easier to avoid Via in the Intel world!
 
Shame is Via is STILL producing crap hardware years later..

The thing is, all those reports over the years, for the most part are
problems I never saw using same chipsets. I have to believe they
didn't know how to set up a system properly or had other crap
hardware/drivers the majority of the time. Is Via blameless?
Sometimes yes, sometimes no, but other products aren't bug-free
either. A LARGE part of the problems surround the price of the
system, the amount of design and testing that went into them... if
Intel chipsets were put on rock-bottom priced boards you'd likely see
far more users bashing Intel chipsets too.

Some people simply should not be building their own PCs, they are lost
when the build goes beyond putting-the-card-in-the-slot, turning the
box on, and clicking the pretty buttons.

Above comment was not addressed towards you, but in general is the
trend I see.


Dave
 
Todd said:
You'd lose that bet I'm afraid.

The 600 was on a Ali Aladdin V AGP set on a Gigabyte GA-5AX.

The 800 was a Via KT133 chipset, which at the time was apparently
better refered than the Ali alternative. Abit KT7-RAID motherboard
(but not using RAID).

It's a lot easier to avoid Via in the Intel world!


Oh well, both were crap chipsets (I forgot about Ali)! My point was that AMD
processors aren't the problem, it's the crap chipsets. Now that nvidia (and
some of the SIS chipsets) work well, AMD can be a good option.
 
kony said:
The thing is, all those reports over the years, for the most part are
problems I never saw using same chipsets.

Come on, like all the problems the KT133's had? You never had ANY problems
with them?
Some people simply should not be building their own PCs, they are lost
when the build goes beyond putting-the-card-in-the-slot, turning the
box on, and clicking the pretty buttons.

But should you buy hardware, try the drivers that come with it and find it
doesn't work right and then have to wait a few months for Via to write
drivers that do work before a system is stable? How many times has Via
released garbage drivers that cause all sorts of flakey problems? Again I'm
sick of paying to be a beta tester and having to work through all these
problems when I can buy another board and "putting-the-card-in-the-slot,
turning the box on" is all I have to do?
 
Come on, like all the problems the KT133's had? You never had ANY problems
with them?

I wasn't using the defective SB cards so my exposure to the risk was
far less, but the KT133 falls under "for the most part", that I do
accept that the KT133 was paired with a southbridge that had PCI
problems in some situations, but in others it was just lower
performance. Having written that, I have set up KT133 based systems
and still regularly communicate with people who I've sold KT133 based
system, who use them every day without issue.
But should you buy hardware, try the drivers that come with it and find it
doesn't work right and then have to wait a few months for Via to write
drivers that do work before a system is stable? How many times has Via
released garbage drivers that cause all sorts of flakey problems? Again I'm
sick of paying to be a beta tester and having to work through all these
problems when I can buy another board and "putting-the-card-in-the-slot,
turning the box on" is all I have to do?

But it isn't that simple... Other Via chipsets SINCE the KT133A have
been no problem for me and many others, they were
"putting-the-card-in-the-slot, turning the box on" easy.
On the other hand I remember a few people bashig the SIS chipset
when their K7S5A boards were flaky, having nothing to do with the Sis
chipset, or various problems with nForce boards, including memory bugs
in nForce2 chipsets. We could point the finger at Via and ignore
other chipset problems but that doesn't assure a trouble-free
motherboard. Motherboards have to be considered on a case-by-case
basis. If you don't want a board with a Via chipset that's fine, but
your conclusion that it will be a problem to use one isn't usually
true.

It is possible Via boards have a higher problem rate but it's also
possible that these Via boards are being bought by a larger percentage
of do-it-yourselfers, who are trying to set up a system themselves
while the Intel CPU based "competition" boards use a far smaller
percentage of Via chipsets, are set up by professional OEM systems
integrators, a bit more likely to know what they're doing than the
avg. Joe would. This would tend to make people having non-Via boards
see fewer problems, but for different reasons. I don't think it's all
that surprising that someone might want to use the newest drivers for
a Via based board or any other manufacturer's product, but considering
this I do think Intel takes 1st place for doing the most testing
before releasing a component product.

As for the RAID controller, since there appears to be a large benefit
to choosing some other controller it may be good to avoid it, but
there isn't this clear disparity with Via chipsets in general beyond
the now-retired for years, 686 southbridge. Sometimes they are lower
performance but at lower cost, an acceptible tradeoff as with any
other hardware... is why we have benchmarks and reviews. Plenty of
reviewers set up and test Via-based boards without these problems you
imply.


Dave
 
stacey said:
Oh well, both were crap chipsets (I forgot about Ali)! My point was
that AMD processors aren't the problem, it's the crap chipsets. Now
that nvidia (and some of the SIS chipsets) work well, AMD can be a
good option.

But, appreciate my perspective--this is somewhat like saying the Yugo
wasn't a bad automobile because...oh, let's say it had a top notch
cylinder heads or something.

It seemed at least for that generation of AMD processors, no matter
how great the processor was, neither of the two competing chipsets
were worth a shit. Why? Well, dunno--you could conjecture...maybe
AMD just didn't get the processor to the chipset makers early enough
for testing, maybe the AMD isn't powerful enough to sway the chipset
makers the way they needed to, or a lot of other reasons that might
have to do with their inferior marketshare to Intel, and that they
don't seem to hold themselves to being able to make a good motherboard
or chipset themselves, so they have to rely on 3rd party ninnies for a
whole solution.

These were the first times I gambled on what I view as the underdog
CPU maker, and I got burnt despite spending a lot of time in careful
component selection. And it doesn't help that I've never had such
problems with any Asus/Intel or intel/intel mobo/processor combo I've
put together. So I can't help but be left with the perception that
it's hard to build a stable AMD based machine.

And don't get me wrong, I really really wanted my AMD machines to work
and be stable and wonderful to cheer the underdog who may well have
superior processor architecture...but alas my fears of stability of
the overall solution turned out to be valid. :-\

Best Regards,
 
Todd said:
or a lot of other reasons that might
have to do with their inferior marketshare to Intel, and that they
don't seem to hold themselves to being able to make a good motherboard
or chipset themselves, so they have to rely on 3rd party ninnies for a
whole solution.

Bingo and I don't blame you. I was "intel only" until very recently and if I
had been burned by those crap chipsets (remember Via made some crap intel
chipsets at that time as well!) I'd still be pro intel. Now I will use
either as long as I don't have to use a Via chipset!

These were the first times I gambled on what I view as the underdog
CPU maker, and I got burnt despite spending a lot of time in careful
component selection. And it doesn't help that I've never had such
problems with any Asus/Intel or intel/intel mobo/processor combo I've
put together. So I can't help but be left with the perception that
it's hard to build a stable AMD based machine.


Yep and I feel the same way about building a stable machine using a Via
chipset when some people say they work great? I've seen too many that
didn't!

And yes AMD should have made their own boards/chipsets and this whole
"unstable" thing =IS= their fault for no doing it themselves.
 
kony said:
Motherboards have to be considered on a case-by-case
basis. If you don't want a board with a Via chipset that's fine, but
your conclusion that it will be a problem to use one isn't usually
true.

It's =more likely= than using any other brand chipset looking at their past
and present performance as a gauge.
but considering
this I do think Intel takes 1st place for doing the most testing
before releasing a component product.

Bingo
 
So I can't help but be left with the perception that
it's hard to build a stable AMD based machine.

.... blaming an entire assembly instead of the specific part, a part no
longer used today. With that attitude the next time your (intel)
based system has a bad part in it, you'll be swearing off both Intel
and AMD, and so on, till an abacus is all that's left?


Millons of AMD based systems work fine. "Sometimes" they're more
trouble, most times they're better performance per $. It's your
choice what to buy, but at least be clear about why you made that
choice, not using superstition to guide you.



Dave
 
kony said:
... blaming an entire assembly instead of the specific part, a part no
longer used today. With that attitude the next time your (intel)
based system has a bad part in it, you'll be swearing off both Intel
and AMD, and so on, till an abacus is all that's left?


Millons of AMD based systems work fine.

And of those, the only two I've had have been a complete pain in my
ass. The rest of the millions won't sway my own experience I'm
afraid.
"Sometimes" they're more trouble, most times they're better
performance per $. It's your choice what to buy, but at least be
clear about why you made that choice, not using superstition to
guide you.

I have designed integrated circuits on the transistor level for the
high end workstation market. I hold a patent in fast static RAM
memory design. Indeed, there are good parts, and there are bad parts.
And one of the things you learn in chip design and system design is
the value of 6-sigma quality in systems with so many piece parts, and
the value of system and individual part qualification cycles...and it
takes a shitload of work. Hell, the same company collaborating
between different product groups to get a system stable has a hard
enough time, but if you separate that task where entirely different
companies that may not be on the same continent are concerned, the
design collaboration becomes a real bitch. AMD doesn't even hold
itself to make its own chipsets and motherboads, so they can always
point the finger at the 3rd party. Pity, there's no way to avoid
using a 3rd party chipset with AMD, though.

Of the millions of AMD systems that you say have worked wonderfully,
I'm afraid I just can't shake the statistic (not superstition) that
100% of the AMD based systems I've had in my posession since the dawn
of the PC revolution have been unstable for me, whilst the Intel-based
systems I've had were stable as can be until they were unusably slow
for modern software.

I'm afraid all the other folks glowing experience with AMD won't sway
me to flying the AMD flag again any time soon.

Your mileage, of course, may vary.

Best Regards,
 
Back
Top