A computer components & hardware forum. HardwareBanter

If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below.

Go Back   Home » HardwareBanter forum » Processors » Overclocking
Site Map Home Register Authors List Search Today's Posts Mark Forums Read Web Partners

Motherboard for P4.



 
 
Thread Tools Display Modes
  #41  
Old July 31st 04, 04:31 AM
Dennis E Strausser Jr
external usenet poster
 
Posts: n/a
Default

"Johannes H Andersen" wrote
in message news:410AF477.32FAE761@nsuvuooiaiosizefitterwiuove swernuaz.com...


JK wrote:

The dual channel of the P4 is relatively slow though since the
memory controller is not on the chip. Look at the actual benchmarks
comparing the two. The article has benchmarks for both socket
754 and socket 939 Athlon 64 chips.

http://www.anandtech.com/cpuchipsets...spx?i=2065&p=1


So?

The dual channel memory 2xDDR400 for the P4 matches the fsb = 800 MHz,
this is the optimal situation.

The P4 is not 800 fsb, it's 200
It works something like the Athlon Xp, but does it twice.
To better explain this, here's an email I got from AMD

Hello Dennis,

Thank you for contacting AMD's Technical Service Center.

The Athlon XP has a Front Side Bus (FSB) that operates at either 266,
333, or 400Mhz. While the physical signal is 133, 166, or 200Mhz, data
is transferred on both the rising and falling edges of the clock
signal. This effectively doubles the data throughput. This is similar
to the operation of DDR memory and 2X AGP. Motherboards that support a
400, 333, 266, and 200MHz front-side bus (FSB) will typically have a
factory-default FSB setting of 200MHz (100MHz system clock) to protect
200MHz FSB processors from accidentally being overclocked. If an Athlon
XP processor, which supports a 400, 333 or 266MHz FSB, is installed on
a motherboard that is configured to operate the FSB at 200MHz, it will
operate at a lower frequency. This is a result of the processor's
multiplier. The function of the multiplier is to multiply the bus
frequency to derive the processor operating frequency.

The actual setting of the FSB may be controlled by the motherboard BIOS
or by a hardware jumper on the motherboard itself. Please consult your
motherboard manufacturer directly to determine how to correctly set the
FSB for your motherboard.

Hope this helps. If you have any other questions, please feel free to
contact me.

Regards,
Jesus
Customer Support Analyst
AMD TSC


We welcome your feedback and suggestions to help us improve our
services to you. To provide this information to us, we ask that you
please click on this link, or copy/paste into your browser, and
complete our short survey. Thanks, in advance, for your comments. Click
here (link
http://asksurvey.amd.com/servicesoft...@comc ast.net).


P.S. Please visit our online technical support tools, Ask AMD
(http://ask.amd.com) and our Processor Support Forums
(http://forums.amd.com). Ask AMD is our online knowledgebase that
contains many solutions to common questions. Our Processor Support
Forums are an online community where users can assist each other with
many different issues. There's a good chance these tools can help
answer your next question!


Original Message Follows:
------------------------


Form Message

Processor Type: Athlon XP
Escalated From: startup
Processor Model: 2200+
Knowledge Base: Processor
Email Address:
Full Name: Dennis E Strausser Jr
Message Body: (FSB) & Bus Speed. If a CPU has a FSB of 133, that would
mean the Bus Speed is 266? So like My 2200+ I have is 266 Bus Speed, and
133 (FSB)? And a 2800+ would be 166 FSB (166.5)? What I need is a link
if I'm right, if I'm wrong, I still want that link. I was trying to tell
some1 that I thought that's the way it works, and he said I'm wrong. So
I'm sure you guys can tell me, after all, they are your CPU's 2200+?
2600 @ 2.17 GHz,,, But you didn't need to know I'm Overclocking one of
your CPU's. Denny. :-)
Subject: FSB & Bus Speed
User Type: Reseller/System Builder
Knowledge Job Ticket: 0000000000169818213:5486
Knowledge Session Log URL:
http://139.95.253.213:80/SRVS/CGI-BI...g,e=0000000000
169818213,K=5486
Location: USA/Canada

What this all means?
I'll break it down.
the rising and falling edges of the clock
signal are still working for the most part, the same way as an amd.
But.
200 x 200 rising and falling edges of the clock signal.
200 x 200 rising and falling edges of the clock signal.
Don't take my word for it, this is just a guess.
But it does seem like a good guess to say that's how it's Hyperthreading
works.
Back when it was just Hyperpipline, I think it helped to keep the cpu
running smooth.
And keep the bottle neck as low as possible.

If anyone has more input on this, or if I'm wrong, correct me.
thx..
Denny. :-)


  #42  
Old July 31st 04, 08:56 AM
David Maynard
external usenet poster
 
Posts: n/a
Default

~misfit~ wrote:

Dave C. wrote:

"maggot" wrote in message
. ..

On Fri, 30 Jul 2004 15:38:28 -0400, "Dave C." wrote:



Why? You got something against incredibly high 3DMark scores in an
extremely stable system? -Dave


Bad history with VIA, That's how you lose customers. Anandtech
recommends the Epox 8KDA3+ which is Nforce3 250gb chipset.


And that's not a bad recommendation. It doesn't necessarily mean
that one is better than the other, though. I see outrageous claims
all the time like "AMD SUCKS" or "AMD ROCKS" or "VIA SUCKS". Claims
like that are clearly based on ignorance. For the most part, AMD is
usually a better deal than Intel, but I'm open-minded enough to build
Intel if it's a better value for what I'm trying to accomplish. Same
with via . . . I know there are better chipsets available, but that
doesn't necessarily mean there is anything wrong with via. My own
primary system is an nforce2, but I just built a via chipset system
for a friend of mine, and I'm tempted to steal it. It's THAT good.
Nobody is doing anybody any favors by steering them away from via.
Any system you build with via will work great, as long as you don't
make the common errors that many builders make, such as cheap power
supply, no-name RAM, etc. In short, if you don't like via, learn to
build a good system. If you don't learn to build a good system, don't
blame the chipset (or the CPU or the video card or the ???) for your
problems. -Dave



How utterly arrogant of you. I, like maggot, dislike VIA chipsets. I hold
this opinion through several years experience of building quite a few
computers, for myself, family, friends and friends of family/friends. I have
used chipsets from Intel, VIA, nVidia, SIS and Ali, probably more too. I
*know* how to build a good system (I don't want them coming back with
problems if I can aviod it) and one of the first rules of building a good
system for performance is 'Don't use a VIA chipset board'.

Oh sure, you can get them running stably and reliably if you know what
you're doing but you won't get the full potential performance out of your
CPU/peripherals. I keep a book detailing all the builds I do and benchmark
results from a battery of benchmarks for each build. I have used the same
CPU in several different mobo's. For a while I went through a
'data-gathering' phase, building and re-building systems with differing
combinations of hardware. I have never seen a VIA chipset board out-perform
a non-VIA board. Ever. In fact the general trend, going from my data
gathered over several years and close to a hundred systems is that VIA
boards tend to run between 5% and 20% slower than competing boards. I no
longer build systems on VIA boards unless the person I'm building for
absolutely insists or comes to me with the parts.

Hardware is my hobby and my passion. I overclock all my own systems. Here's
an example for you from my records:

Celeron Tualatin 1.4Ghz on Gigabyte/VIA board:

100FSB (1.4Ghz) 87.3 CPU Mark 99 marks
110FSB (1.54Ghz) 98.6 marks
115FSB (1.61Ghz) 105 marks.

Same CPU in an MSI/Intel 440BX board (with adapter):

100FSB 110 marks
110FSB 118 marks
115FSB 127 marks

(All settings tested extensively with Prime95 for stability)


Well, not all VIA chipsets have the same performance so I don't think it's
quite fair to base a generic condemnation on one version.

I still think the major difference is that VIA mobo has built in video with
it's frame buffer taken from main memory and it's sucking up memory
bandwidth with constant display refresh.

Having said that, I think even the 'best' of the socket 370 VIA SDR SDRAM
chipsets lagged behind the BX by a couple % in memory bandwidth, but it was
closer than the others.


That last build is how I left it, it's one of my SETI boxes, been running
perfectly 24/7 for a while now. Incidently it benchmarks almost identically
to an XP2000+/VIA chipset build I did (under duress) a short while ago.

Wanna buy the VIA board? It's sitting on a shelf, along with three other VIA
boards. One of which was in own system, I bought it at a PC builder's
liquidation auction, which I replaced with an nForce board. (I wonder if the
fact that they used VIA boards almost exclusively had anything to do with
them going into liquidation?)
--
~misfit~



  #43  
Old July 31st 04, 11:11 AM
~misfit~
external usenet poster
 
Posts: n/a
Default

David Maynard wrote:
~misfit~ wrote:

Dave C. wrote:

"maggot" wrote in message
...

On Fri, 30 Jul 2004 15:38:28 -0400, "Dave C."
wrote:



Why? You got something against incredibly high 3DMark scores in
an extremely stable system? -Dave


Bad history with VIA, That's how you lose customers. Anandtech
recommends the Epox 8KDA3+ which is Nforce3 250gb chipset.

And that's not a bad recommendation. It doesn't necessarily mean
that one is better than the other, though. I see outrageous claims
all the time like "AMD SUCKS" or "AMD ROCKS" or "VIA SUCKS". Claims
like that are clearly based on ignorance. For the most part, AMD is
usually a better deal than Intel, but I'm open-minded enough to
build Intel if it's a better value for what I'm trying to
accomplish. Same with via . . . I know there are better chipsets
available, but that doesn't necessarily mean there is anything
wrong with via. My own primary system is an nforce2, but I just
built a via chipset system for a friend of mine, and I'm tempted to
steal it. It's THAT good. Nobody is doing anybody any favors by
steering them away from via. Any system you build with via will
work great, as long as you don't make the common errors that many
builders make, such as cheap power supply, no-name RAM, etc. In
short, if you don't like via, learn to build a good system. If you
don't learn to build a good system, don't blame the chipset (or the
CPU or the video card or the ???) for your problems. -Dave



How utterly arrogant of you. I, like maggot, dislike VIA chipsets. I
hold this opinion through several years experience of building quite
a few computers, for myself, family, friends and friends of
family/friends. I have used chipsets from Intel, VIA, nVidia, SIS
and Ali, probably more too. I *know* how to build a good system (I
don't want them coming back with problems if I can aviod it) and one
of the first rules of building a good system for performance is
'Don't use a VIA chipset board'.

Oh sure, you can get them running stably and reliably if you know
what you're doing but you won't get the full potential performance
out of your CPU/peripherals. I keep a book detailing all the builds
I do and benchmark results from a battery of benchmarks for each
build. I have used the same CPU in several different mobo's. For a
while I went through a 'data-gathering' phase, building and
re-building systems with differing combinations of hardware. I have
never seen a VIA chipset board out-perform a non-VIA board. Ever. In
fact the general trend, going from my data gathered over several
years and close to a hundred systems is that VIA boards tend to run
between 5% and 20% slower than competing boards. I no longer build
systems on VIA boards unless the person I'm building for absolutely
insists or comes to me with the parts.

Hardware is my hobby and my passion. I overclock all my own systems.
Here's an example for you from my records:

Celeron Tualatin 1.4Ghz on Gigabyte/VIA board:

100FSB (1.4Ghz) 87.3 CPU Mark 99 marks
110FSB (1.54Ghz) 98.6 marks
115FSB (1.61Ghz) 105 marks.

Same CPU in an MSI/Intel 440BX board (with adapter):

100FSB 110 marks
110FSB 118 marks
115FSB 127 marks

(All settings tested extensively with Prime95 for stability)


Well, not all VIA chipsets have the same performance so I don't think
it's quite fair to base a generic condemnation on one version.


You're right. However, that's just the most recent and best documented
example I have. The fiancee's niece got into my previous notebook a year or
so ago and destroyed it.

I still think the major difference is that VIA mobo has built in
video with it's frame buffer taken from main memory and it's sucking
up memory bandwidth with constant display refresh.


In the case of that mobo you're right.

Having said that, I think even the 'best' of the socket 370 VIA SDR
SDRAM chipsets lagged behind the BX by a couple % in memory
bandwidth, but it was closer than the others.


As I said, I've never met a VIA board I've been happy with. Maybe I've just
missed the good ones?
--
~misfit~


  #44  
Old July 31st 04, 11:11 AM
Johannes H Andersen
external usenet poster
 
Posts: n/a
Default



Dennis E Strausser Jr wrote:

"Johannes H Andersen" wrote
in message news:410AF477.32FAE761@nsuvuooiaiosizefitterwiuove swernuaz.com...


JK wrote:

The dual channel of the P4 is relatively slow though since the
memory controller is not on the chip. Look at the actual benchmarks
comparing the two. The article has benchmarks for both socket
754 and socket 939 Athlon 64 chips.

http://www.anandtech.com/cpuchipsets...spx?i=2065&p=1


So?

The dual channel memory 2xDDR400 for the P4 matches the fsb = 800 MHz,
this is the optimal situation.

The P4 is not 800 fsb, it's 200
It works something like the Athlon Xp, but does it twice.


Yes, yes but...

The P4/800 still matches dual channel memory 2xDDR400 optimally with a max
memory bandwidth of 6.4 GB/s. The AMD socket 754 doesn't.

Whether you then call it 800 fsb or 200 fsb is a matter of semantics. The
clock multiplier is indeed applied to 200 MHz, but dual channel DDR gives
you the 800 MHz data rate.
  #45  
Old July 31st 04, 01:32 PM
Courseyauto
external usenet poster
 
Posts: n/a
Default

JK wrote:

The dual channel of the P4 is relatively slow though since the
memory controller is not on the chip. Look at the actual benchmarks
comparing the two. The article has benchmarks for both socket
754 and socket 939 Athlon 64 chips.

http://www.anandtech.com/cpuchipsets...spx?i=2065&p=1



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

If you throw out the FX 55 which is not even available yet,the intel CPU's show
up quite a bit better in the benchmarks. DOUG
  #46  
Old July 31st 04, 01:46 PM
Dave C.
external usenet poster
 
Posts: n/a
Default


How utterly arrogant of you. I, like maggot, dislike VIA chipsets. I hold
this opinion through several years experience of building quite a few
computers, for myself, family, friends and friends of family/friends. I

have
used chipsets from Intel, VIA, nVidia, SIS and Ali, probably more too. I
*know* how to build a good system (I don't want them coming back with
problems if I can aviod it) and one of the first rules of building a good
system for performance is 'Don't use a VIA chipset board'.

Oh sure, you can get them running stably and reliably if you know what
you're doing but you won't get the full potential performance out of your
CPU/peripherals.


(snip)

I dare to post facts rather than opinions, and that makes me arrogant
somehow? There are several chipset makers, and all of them (specifically
including VIA) have had the performance edge at various times until the
competition fired their next volley. To make a blanket statement like "
'one of the first rules of building a good system for performance is 'Don't
use a VIA chipset board' " is just plain stupid.

In fact, until the nforce3 250 went live, the VIA K8T800 was arguably the
best choice for Athlon64. Even now, the performance numbers between VIA,
SIS and nvidia are virtually identical, if you are building on the Athlon64
platform. That would make VIA a wise choice based on selection of
mainboards and price, though you'll find some good nforce3 250 boards
competitively priced, also.

If you are building a P4, the best value in chipsets at the moment would be
SIS, VIA and Intel in that order (though not too many boards use the SIS
655TX). While the three chipsets have performance numbers that are
virtually identical, each has its own price point and its own strengths and
weaknesses. If you want a good gaming system that won't cost an arm and a
leg, VIA PT880 is a good choice for the P4 at the moment. The Intel 875P
will offer virtually identical performance to the VIA PT880, but the 875P
also costs more. The 655TX is a little faster than VIA PT880, but the
selection of those boards is somewhat limited. All things considered, VIA
PT880 would be the best choice for many P4 builders, at the moment.

Note I said 'at the moment' as the technology is constantly changing. I
don't doubt that you benchmarked a celeron on a Intel BX board faster than a
similar VIA chipset board. AT THAT MOMENT, the Intel chipset was clearly
better. You are doing nobody any favors (least of all yourself) by
automatically dismissing any motherboard with a via chipset. Depending on
when you build, VIA might be the best choice. Yes, for performance,
lso. -Dave



  #47  
Old July 31st 04, 01:46 PM
Dave C.
external usenet poster
 
Posts: n/a
Default


As I said, I've never met a VIA board I've been happy with. Maybe I've

just
missed the good ones?
--
~misfit~


Apparently so. -Dave


  #48  
Old July 31st 04, 01:49 PM
Dave C.
external usenet poster
 
Posts: n/a
Default

You are now in my killfile.

ROTFLMAO

I expect I'll be joining Dave soon. I just hope my poxy AMD system
(NF2) doesn't bomb out on me before I get to see the post. Of course
that's assuming the AMD + VIA system on the other side of the desk
doesn't blow up and kill everyone in the house first.


Tim


Funny. -Dave


  #49  
Old July 31st 04, 01:56 PM
Dave C.
external usenet poster
 
Posts: n/a
Default

"Dave C." wrote:


VIA is often very unstable. Very bad reputation.


AMD is often very unstable.


Do you have any statistical evidence of that? I doubt that you
could provide any.One can build an ustable system with an Intel
processor or with an AMD processor if they don't know what they
are doing.


Whoa. I see that one went right over your head. Someone posted "VIA is
often very unstable. Very bad reputation." I responded with "AMD is often
very unstable. Very bad reputation." I did it to illustrate how stupid the
other poster sounded when he was bashing VIA with inaccurate
generalizations. I build with all chipsets using both AMD and Intel
processors. I happen to prefer nvidia chipsets with AMD processors at the
moment, but have no strong feelings against any other combination you could
care to name. In fact, my last build was a VIA chipset Intel system that
KICKS ASS, to put it bluntly. -Dave


  #50  
Old July 31st 04, 03:31 PM
David Maynard
external usenet poster
 
Posts: n/a
Default

~misfit~ wrote:

David Maynard wrote:

~misfit~ wrote:


Dave C. wrote:


"maggot" wrote in message
m...


On Fri, 30 Jul 2004 15:38:28 -0400, "Dave C."
wrote:




Why? You got something against incredibly high 3DMark scores in
an extremely stable system? -Dave


Bad history with VIA, That's how you lose customers. Anandtech
recommends the Epox 8KDA3+ which is Nforce3 250gb chipset.

And that's not a bad recommendation. It doesn't necessarily mean
that one is better than the other, though. I see outrageous claims
all the time like "AMD SUCKS" or "AMD ROCKS" or "VIA SUCKS". Claims
like that are clearly based on ignorance. For the most part, AMD is
usually a better deal than Intel, but I'm open-minded enough to
build Intel if it's a better value for what I'm trying to
accomplish. Same with via . . . I know there are better chipsets
available, but that doesn't necessarily mean there is anything
wrong with via. My own primary system is an nforce2, but I just
built a via chipset system for a friend of mine, and I'm tempted to
steal it. It's THAT good. Nobody is doing anybody any favors by
steering them away from via. Any system you build with via will
work great, as long as you don't make the common errors that many
builders make, such as cheap power supply, no-name RAM, etc. In
short, if you don't like via, learn to build a good system. If you
don't learn to build a good system, don't blame the chipset (or the
CPU or the video card or the ???) for your problems. -Dave


How utterly arrogant of you. I, like maggot, dislike VIA chipsets. I
hold this opinion through several years experience of building quite
a few computers, for myself, family, friends and friends of
family/friends. I have used chipsets from Intel, VIA, nVidia, SIS
and Ali, probably more too. I *know* how to build a good system (I
don't want them coming back with problems if I can aviod it) and one
of the first rules of building a good system for performance is
'Don't use a VIA chipset board'.

Oh sure, you can get them running stably and reliably if you know
what you're doing but you won't get the full potential performance
out of your CPU/peripherals. I keep a book detailing all the builds
I do and benchmark results from a battery of benchmarks for each
build. I have used the same CPU in several different mobo's. For a
while I went through a 'data-gathering' phase, building and
re-building systems with differing combinations of hardware. I have
never seen a VIA chipset board out-perform a non-VIA board. Ever. In
fact the general trend, going from my data gathered over several
years and close to a hundred systems is that VIA boards tend to run
between 5% and 20% slower than competing boards. I no longer build
systems on VIA boards unless the person I'm building for absolutely
insists or comes to me with the parts.

Hardware is my hobby and my passion. I overclock all my own systems.
Here's an example for you from my records:

Celeron Tualatin 1.4Ghz on Gigabyte/VIA board:

100FSB (1.4Ghz) 87.3 CPU Mark 99 marks
110FSB (1.54Ghz) 98.6 marks
115FSB (1.61Ghz) 105 marks.

Same CPU in an MSI/Intel 440BX board (with adapter):

100FSB 110 marks
110FSB 118 marks
115FSB 127 marks

(All settings tested extensively with Prime95 for stability)


Well, not all VIA chipsets have the same performance so I don't think
it's quite fair to base a generic condemnation on one version.



You're right. However, that's just the most recent and best documented
example I have. The fiancee's niece got into my previous notebook a year or
so ago and destroyed it.


I still think the major difference is that VIA mobo has built in
video with it's frame buffer taken from main memory and it's sucking
up memory bandwidth with constant display refresh.



In the case of that mobo you're right.


Having said that, I think even the 'best' of the socket 370 VIA SDR
SDRAM chipsets lagged behind the BX by a couple % in memory
bandwidth, but it was closer than the others.



As I said, I've never met a VIA board I've been happy with. Maybe I've just
missed the good ones?


Oh, lordy. I sure as heck ain't gonna say anything is 'good' at THIS stage
of the conversation. LOL


 




Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
A7N8X Motherboard Low Temperature Sensitivity, CMOS Checksum Error kony General 6 October 18th 04 05:38 AM
AGP Voltage problem John Overclocking 4 February 29th 04 11:51 AM
AOPEN AX4BS-V MOTHERBOARD - won't boot - won't POST - help please!! lucy General 3 November 29th 03 06:21 AM
I need to replace a bad Asus A7V motherboard David Cook General 1 October 31st 03 02:40 AM
your motherboard redbuffalo003 General 0 September 21st 03 04:08 PM


All times are GMT +1. The time now is 06:40 PM.


Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Copyright ©2004-2024 HardwareBanter.
The comments are property of their posters.