overclocker

  • Thread starter Thread starter P J C
  • Start date Start date
P

P J C

Hey all, I just got my 9800pro sapphire, which I love.
I am thinking about overclocking, but I am new to the ati club, was an
nvidia fanboy for a bit there.
What tool do you all recommend and also, what core / mem speeds do you think
are achievable.
thanks
 
P J C said:
Hey all, I just got my 9800pro sapphire, which I love.
I am thinking about overclocking, but I am new to the ati club, was an
nvidia fanboy for a bit there.
What tool do you all recommend and also, what core / mem speeds do you think
are achievable.
thanks
Click to expand...
I'm also new to ATI coming over from years with Nvidia. I use RivaTuner it
works great. I also have tried AtiTool, which also works fine. AtiTool is
handy because it lets you test for artifacts. It has a little 3D test window
that works great.
Got to tell you I'm not all that excited about what my card can do. I have
the same card as you. Have you looked at your memory modules to see what
brand and speed they are? I have Samsung 3.25 on mine. I tested my card for
the best overclock for several days and what I ended up with is 412/376. If
I lower the core to 400 I can up the ram to 380. But I get better
performance with the higher core setting. But after throwing ever benchmark
and real game testing my results are not worth it. I only gain 1-2 FPS no
matter what it is. Now someone posted here last week that that's OK because
what the clocking is doing is helping the card from dipping when there is
heavy load on it. I'm really not sure if that true or not. When I have my
Ti-4200 I could clock it to 310/660
I would see at lest a 6 FPS increase and in some games even 10FPS. So I'm
used to really seeing a big improvement with clocking. But all said, the
card is plenty fast at stock speed. What you need to do to really see the
improvement over your old card is to crank up the AA and Anisotropic
Filtering. Man does it make a huge improvement. In some of my games the
different is really amazing. The 9800Pro just plan kicks major eyecandy butt
over what Nvidia has to offer. Post back and let me know what you can get
out of your card OK?
JLC
 
Sounds like you may be cpu bottlenecked. My 9800np clcoked to pro speeds
gives a good increase.

Mike
 
try the omega drivers from www.omegacorner.com
which include a radclocker for overclocking and artifact tester too. try the
newest drivers by all means but get the 3.7s as a backup - most
stable/compatible drivers there are.
 
Mike P said:
Sounds like you may be cpu bottlenecked. My 9800np clcoked to pro speeds
gives a good increase.
Click to expand...
I'd like to see some data on how much of incresce in FPS (not benchmark
scores) there is with the same card overclocked at the same speed, but with
different CPU's. I have a P4 2Ghz clocked to 2.2. Anyone out there that has
say a 3Ghz that can inform me on whay there getting. JLC
 
Hey JLC,
thanks for the great response.
I tried the ATITool, seems very elaborate, very well done. Honestly,
I'm happy at stock speeds and it takes waaaay too long to figure out the max
core and mem spds.
 
It isn't exact but the closest thing I know of to that is the comparison
database for 3dmark2001. If I ever get the energy maybe I'll drop my xp2500
down to stock speed and do some benches, then bench again at xp3200
speeds... I'll post here if I do.

Mike
 
JLC said:
I'd like to see some data on how much of incresce in FPS (not benchmark
scores) there is with the same card overclocked at the same speed, but with
different CPU's. I have a P4 2Ghz clocked to 2.2. Anyone out there that has
say a 3Ghz that can inform me on whay there getting. JLC
Click to expand...


I'm not about to overclock my 9800 Pro, but I have a 2.8GHz 800MHz FSB
Pentium 4 and I recall getting about 15,700 in 3DMark2001SE at default
settings and no overclocking. How does that compare to your system? It may
help give an idea as to bottlenecks.
 
NightSky 421 said:
I'm not about to overclock my 9800 Pro, but I have a 2.8GHz 800MHz FSB
Pentium 4 and I recall getting about 15,700 in 3DMark2001SE at default
settings and no overclocking. How does that compare to your system? It may
help give an idea as to bottlenecks.
Click to expand...
What I realize now after doing lots of testing is that yes, if you run a BM
like 3DMark your score is going to look impressive when you OC, but if you
look at how much theactually FPS increase it's not very much. As I said I'm
interested in real game testing.
I've been playing around with AquaMark which is suppose to represent real
gaming vs. a BM score. It sure looks cool!
 
newest drivers by all means but get the 3.7s as a backup - most
stable/compatible drivers there are.
Click to expand...

Everyone just loves 3.7. Don't know why. I think 4.2 is also just as good
and fast as 3.7. That's on an old r200 8500le card.

I think 3.7 are popular due to the "special magical numbers" 3 and 7...
lucky lucky
 
JLC said:
What I realize now after doing lots of testing is that yes, if you run a BM
like 3DMark your score is going to look impressive when you OC, but if you
look at how much theactually FPS increase it's not very much. As I said I'm
interested in real game testing.
Click to expand...


Fair enough, I agree real gaming performance is the most important.

I've been playing around with AquaMark which is suppose to represent real
gaming vs. a BM score. It sure looks cool!
Click to expand...


I installed the AquaMark benchmark myself and it looks great!
 
NightSky 421 said:
Fair enough, I agree real gaming performance is the most important.




I installed the AquaMark benchmark myself and it looks great!
Click to expand...


What I used to do when I cared about 3D Mark, was use demos recorded in
games I actually played to get a better assessment of drivers, as I found
sometimes the drivers that gave me the best 3D Mark performance game me the
worst real game performance.
 
Back
Top