Running Barton 2500+ at 200FSB, but not overclocking ... on A7N8X

[Harry Muscle]

I was wondering if someone could verify if this kind of a setup is possible:

I'm thinking of getting an Athlon XP Barton 2500+ and run it on a Asus A7N8X
Deluxe board. However, since the board can handle 200MHz FSB, I'm hoping to
be able to run my chip at that FSB speed, but I'm not the overclocking type.
So here's what I wanted to do:

Change the FSB to 195MHz
Set the Multiplier to 9.5
My chip would now be running at 1852.5Mhz as opposed to the origianl 1833.
This I'm comfortable with. Obviously I would require PC3200 memory to do
this.

However, the parts that I'm confused about are ... there was an article on
Tom's Hardware about the Barton core and they mentioned that they tried
running the chips at 200MHz FSB, however, they were unable to get it stable.
From what I understand they changed the FSB as well as the multiplier in
order to not overclock the chip. From the chips point of view, does it
really matter what the FSB and multiplier are set at, as long as the actual
speed is the same?

Also, are Barton chips multiplier unlocked? Or do you have to do that
"connect the two pins" thing in order to choose a different multiplier?

Thanks,
Harry

 

[Big George]

Sorry for thread jumping but also same question but different cpu.

Have a 1800xp , Have managed 166 fsb x 9.0 . = 1577, not much over what a
1800xp should
be running at. 3D Marks improved and sytem was stable.

Cpu jumped from 45c to 49c which is still well under max temps.

Anyone got an 1800xp any higher?

Also running pc3200 (2 x 256mb in dual channel)
A7N8X Deluxe , rev 1.06 , 1004 Bios


George

 

[Ben Pope]

I was wondering if someone could verify if this kind of a setup is
possible:

I'm thinking of getting an Athlon XP Barton 2500+ and run it on a
Asus A7N8X Deluxe board. However, since the board can handle 200MHz
FSB, I'm hoping to be able to run my chip at that FSB speed, but I'm
not the overclocking type. So here's what I wanted to do:

That should be fine.

Change the FSB to 195MHz
Set the Multiplier to 9.5
My chip would now be running at 1852.5Mhz as opposed to the origianl
1833. This I'm comfortable with. Obviously I would require PC3200
memory to do this.

Indeed, but a few MHz is neither here nor there, so you may as well go
200MHz*9.5 - thats 3.7% over spec, which considering people usually get
2.2GHz (20%) without problems is not a lot.

192MHz*9.5 is closest to original spec without overclocking anything.

However, the parts that I'm confused about are ... there was an
article on Tom's Hardware about the Barton core and they mentioned
that they tried running the chips at 200MHz FSB, however, they were
unable to get it stable. From what I understand they changed the FSB
as well as the multiplier in order to not overclock the chip. From
the chips point of view, does it really matter what the FSB and
multiplier are set at, as long as the actual speed is the same?

Doesn't matter as far as the chip is concerned, final MHz is final MHz. The
stability problems may have arisen from the board/chipset.

Also, are Barton chips multiplier unlocked? Or do you have to do that
"connect the two pins" thing in order to choose a different
multiplier?

All unlocked.

Other should be careful of the board though, if you want to traverse the
12.5, 13 multiplier, it'll need to support the 5th bit, which the A7N8X does
not. This should not be a problem though for what you want to achieve.

Ben

 

[Robert Sudbury]

I am running an A7N8X DX2 UAY (nForce2) board and a UAYZ (nForce2 ULTRA400)
board both with Barton 2500+ and two sticks of OCZ PC3200 rev2 CL2 256MB in
dual DDR mode at 10x200, 100%, 7-3-3-2 ( or 2-3-3-7 take yer pick) happily,
with no voltage changes.

 

[Doug Collette]

You should have no problems doing what you propose and the multipliers
are unlocked - no surgery required. I have a Barton 2500 with an
A7N8X Deluxe Rev. 2 board and am currently running at at 10.5 x 205
MHz and it is completely stable.

I have experimented with the FSB speed and have been able to get it up
to 220 MHz with a multiplier of 10.5, but decided to back it down in
the interest of cooler running and longer life.

Good luck,

Doug

 

[Doug Collette]

You should have no problems doing what you propose and the multipliers
are unlocked - no surgery required. I have a Barton 2500 with an
A7N8X Deluxe Rev. 2 board and am currently running at at 10.5 x 205
MHz and it is completely stable.

I have experimented with the FSB speed and have been able to get it up
to 220 MHz with a multiplier of 10.5, but decided to back it down in
the interest of cooler running and longer life.

Good luck,

Doug

 

[Ted K]

Unless you're very unlucky, that will work just fine.
But, consider going even higher still. If you can get 10 x 200MHz (2GHz) to
work fine, then why not go ahead ? It's free speed - you have nothing to
lose !


Yup.

Can you do 10 X 200MHz with the retail box cooler/fan? If not, how
high can you go?

Also, somewhat unrelated, are software cooling pgms that use the halt
function of any value, as related to the answer to the query above?

 

[JK]

Can you do 10 X 200MHz with the retail box cooler/fan? If not, how
high can you go?

probably yes

People normally do not buy the box cooler. The best air-cooler should
be used, and you can then go to 2200 with standard voltage (1.65), and
2400 with 1.85 volt, aircooled.

The Zalman 7000A-C is a new model compatible with all socketA and 478.

Or the Coolermaster Aero7+

Also, somewhat unrelated, are software cooling pgms that use the halt
function of any value, as related to the answer to the query above?

No, that is cheating.

best regards

John

 

[rstlne]

Also, somewhat unrelated, are software cooling pgms that use the halt

No, that is cheating.

best regards

John

Yea, I wouldt think that using grantstop to keep the temp down would be a
good idea, since it's going to be clock'd outside of spec then the last
thing you'll want is the temp jutting up and down, when you turn your pc off
it cools slower cause the fans stop, so it's not really the same thing.. but
I agree, it'c heating

 

[Ben Pope]

Yea, I wouldt think that using grantstop to keep the temp down would
be a good idea, since it's going to be clock'd outside of spec then
the last thing you'll want is the temp jutting up and down, when you
turn your pc off it cools slower cause the fans stop, so it's not
really the same thing.. but I agree, it'c heating

It certainly does keep the processor cooler, but only when it's not under
full load. Under full load it becomes far less useful and thats when you
need the coolest temps.

Ben

 

[Nom]

Can you do 10 X 200MHz with the retail box cooler/fan?

Try it and see ! If it's not stable, then no - back off. If it is, then
yes - go higher.

If not, how
high can you go?

There are no fixed limits - your highest-point my be completely different to
mine, for example.

Also, somewhat unrelated, are software cooling pgms that use the halt
function of any value, as related to the answer to the query above?

Nope, not in 2000/XP.