To overclock an ATI 9800 Pro I've been told to just run ATI Tools. Ok I
did that but all I got was a default 378 core that couldn't get past
386!!

Do I need to up the voltage on the card first or do I have a cooling
issue?

How do I up the voltage? Via the mobo? My ABit IC7-MAX3 allows me to up
the voltage on the APG (currently set to a default 1.55v)

<(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> To overclock an ATI 9800 Pro I've been told to just run ATI Tools. Ok I
> did that but all I got was a default 378 core that couldn't get past
> 386!!
You should not have to overvolt the card to get a reasonable overclock. My
ATI retail box 9800 Pro R350 128Mb will run at 9800XT spec (412/ 730) on the
stock cooler with no issues. It's been set at this for 18 months via ATI
Tool on bootup. You don't mention what brand of card you have. It's
extremely unusual for a 9800 Pro core to max out at 386Mhz. How and why
won't it go past 386 core? Artifacting? Lockup? What speed is your memory
set at?
You may have a case heat issue or a GPU hs/f issue going on. What's your
case temp and ventilation setup?

<(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> To overclock an ATI 9800 Pro I've been told to just run ATI Tools. Ok I
> did that but all I got was a default 378 core that couldn't get past
> 386!!

What happens if you go past 386? Do visual artifacts appear?

> Do I need to up the voltage on the card first or do I have a cooling
> issue?

Most likely a cooling issue. The stock heat sink is marginal. Better to get
something like an Arctic Cooling VGA Silencer. It cools better, produces
less noise, and should last longer.

Heat increases with the square of voltage. Upping the voltage requires
better cooling first.

> How do I up the voltage? Via the mobo? My ABit IC7-MAX3 allows me to up
> the voltage on the APG (currently set to a default 1.55v)

The AGP voltage is different from the GPU core voltage. You can set the GPU
core voltage either with a hacked card BIOS or a hardware resistor mod. The
latter should only be attempted if you have a multimeter.

--
"War is the continuation of politics by other means.
It can therefore be said that politics is war without
bloodshed while war is politics with bloodshed."

I find it hard to beleive that heat increases with the square of the
voltage. Where did you get this relationship from? What is the exact
formula? Is it just the power formula P = V * (V/R)? This would be a rough
approximation I'd imagine since GPU's aren't mere resistors. I'd imagine at
the switching frequencies of a GPU, inductive loads would be significant.

--
Doug
"First of One" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> <(E-Mail Removed)> wrote in message
> news:(E-Mail Removed)...
>
> Heat increases with the square of voltage. Upping the voltage requires
> better cooling first.
>
>> How do I up the voltage? Via the mobo? My ABit IC7-MAX3 allows me to up
>> the voltage on the APG (currently set to a default 1.55v)
>
> The AGP voltage is different from the GPU core voltage. You can set the
> GPU core voltage either with a hacked card BIOS or a hardware resistor
> mod. The latter should only be attempted if you have a multimeter.
>
> --
> "War is the continuation of politics by other means.
> It can therefore be said that politics is war without
> bloodshed while war is politics with bloodshed."
>
>
>

Yes, it is definitely an approximation, based on P = V * (V/R). Tom's
Hardware once cited the same estimate. It's actually not as bad as you
think. For example, a processor at 1.75 V dissipating 60 W, would dissipate
67 W at 1.85 V, not really a big difference.

Resistive heating blocks are often used as the dummy CPU in heat sink
measurements, so there exists the notion that resistors would be a
half-decent representation of the CPU's thermal output.

--
"War is the continuation of politics by other means.
It can therefore be said that politics is war without
bloodshed while war is politics with bloodshed."

"pigdos" <(E-Mail Removed)> wrote in message
news:euZxf.4163$_(E-Mail Removed)...
>I find it hard to beleive that heat increases with the square of the
>voltage. Where did you get this relationship from? What is the exact
>formula? Is it just the power formula P = V * (V/R)? This would be a rough
>approximation I'd imagine since GPU's aren't mere resistors. I'd imagine at
>the switching frequencies of a GPU, inductive loads would be significant.

I'm currently using a thermaltake VGA Waterblock with a Bigwater SE
setup. I removed it today just to make sure I had a solid contact with
the GPU and I do.

It shows my default Core is 378/338, is that normal for a 9800 Pro?

One other thing. I am currently over clocking my 2.6C intel to 3.25Ghz.
Very stable for a week now. Fully Prime95 tested too.

<(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> I'm currently using a thermaltake VGA Waterblock with a Bigwater SE
> setup. I removed it today just to make sure I had a solid contact with
> the GPU and I do.

That's interesting. This system should be at least as good as the best
air-cooled setups. Is the Radeon placed upstream from the P4? What are the
water temps? I wonder if the hot CPU is heating up the water for the video
card.

> It shows my default Core is 378/338, is that normal for a 9800 Pro?

Yes, it is.

--
"War is the continuation of politics by other means.
It can therefore be said that politics is war without
bloodshed while war is politics with bloodshed."

Your system is going to be bottlenecked by that 9800 Pro in most any modern
FPS (Doom3, Quake IV, HL2, Far Cry). You'll never be able to overclock the
9800 Pro enough to overcome that.

--
Doug
<(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> One other thing. I am currently over clocking my 2.6C intel to 3.25Ghz.
> Very stable for a week now. Fully Prime95 tested too.
>

Good idea the COU feeds the VGA waterblock so maybe your right

I'll have to reverse them.