Radeon 9600 PRO Overclocking

nelg said:
I think some people might owe TSMC an apology. :oops:
I think a lot might be to do with the process chosen - "tried and tested" vs "bleeding edge".
 
RussSchultz said:
Bull crap.
The fab has plenty to do with whether or not a chip works. More than plenty. They provide all the data for the sims you run and if they've got it wrong--then your chip doesn't work as advertized. I worked on a design that spent six months in limbo because the fab wouldn't own up to selling an SRAM cell that was never qualified in production.
I'll agree with that. I've heard tales of processes where, eg, the registers didn't actually work reliably!
 
worm[Futuremark said:
]Hmm.. I noticed that you used Trilinear for Texture Filtering, instead of the default Optimal? By mistake or was there some problems with the Optimal setting?

I'm not too keen on the default settings in 3DMark03 - 'Optimal' seems a bit of a wooley term and you never really know what its using, and the 4X AF is also removed. When I test 3DMark03 I always put it to trilinear and 1x AF for consistency sake and the results were submitted like that to maintain parity with the test results in the review.
 
Shee...

with such an OC-beast at hand, I cannot understand why on earth ATi has not changed the product portfolio toward something like this :

Code:
R9600 :     300MHz Core; 200MHz RAM on R9200-PCB
R9600Pro:   375MHz Core; 275MHz RAM on R9200Pro-PCB
R9600U :    450MHz Core; 350MHz RAM on new PCB

This would have made it possible to cover most lowend-mainstream markets with one chip. The lower end card would be really cheap to produce and would give at least the same performance as the R9200Pro and R8500 but DX9 and the R9600-Ultra would be as fast as an R9500Pro.
So with one chip and 2 old PCB's (+one new) they would be able to cover the market from $99 till $229.

So why hasn't ATi done this? Or is my understanding completely wrong?
 
Dave> When do you think you will be able to try to OC a 5600 U part? :?: Thx!

mboeller> Perhaps it's just a question of yield, which are not as good as anticipated?
 
mboeller said:
This would have made it possible to cover most lowend-mainstream markets with one chip. The lower end card would be really cheap to produce and would give at least the same performance as the R9200Pro and R8500 but DX9 and the R9600-Ultra would be as fast as an R9500Pro.
So with one chip and 2 old PCB's (+one new) they would be able to cover the market from $99 till $229.

So why hasn't ATi done this? Or is my understanding completely wrong?

I agree with you making this point simply because this chip scales so damn well.

I don't know reason why ATI didn't take this route themselves, but the important thing is that nothing should stop a big OEM from putting the chip into three different market segments.

One reason could be that ATI just don't have the need to release the chip below 325 MHz - maybe because the yields are superb there already and because it's cost-efficient to implement this chip in value PCB's.

If that's the case it's primarily up to the OEM's to decide the market segment from chosing memory speed etc.
 
Insane. Brent and the other folks over at hardocp pushed it even further:

Needless to say, we were shocked by the incredible overclocking ability of this graphics card. With a stable core frequency of 567MHz, we were operating the card no less than 167MHz above the default frequency! With a simple movement of the slider within the drivers, the card was suddenly given an almost 50% increase in core frequency.

:arrow: http://www.hardocp.com/article.html?art=NDY0
 
Again, the overclocking of this card is both insane and impressive. Can anyone point me to links where this matters in a game benchmark or two? Sorry for being lazy!
 
Evildeus: I'll try and stick the 5600U board back in the test rig tonight to give it a whirl.

mboeller/LeStoffer: Something I noted in the review is that I think this chip will end up covering the low end, but its likely that prices on TSMC's 130nm process may be prohibiting it right now. In my opinion, as soon as chip prices start falling this will pahse ou the 9000/9200 products quite quickly (especially since I'm hearing they are having issues with 9200 PRO's). If yeilds are good at these high speeds I think ATI would be crazy not to offer a higher spec product though.

Rev: This that what you're after?
 
LeStoffer said:
I don't know reason why ATI didn't take this route themselves, but the important thing is that nothing should stop a big OEM from putting the chip into three different market segments.

One reason could be that ATI just don't have the need to release the chip below 325 MHz - maybe because the yields are superb there already and because it's cost-efficient to implement this chip in value PCB's.

If that's the case it's primarily up to the OEM's to decide the market segment from chosing memory speed etc.

But then ATi wastes money! They could charge more money when would offer an 450MHz part too. The other parts could be sold at the normal price and the highend part at an higher price without any problem. When an OEM does this he can charge more money, but ATi is left in the dust. With the low earnings @ ATi this is imho a big mistake from their side.

I have choosen the 300, 375 and 450MHz core-speeds only to better distinguish the three parts from each other.
 
mboeller said:
Shee...

with such an OC-beast at hand, I cannot understand why on earth ATi has not changed the product portfolio toward something like this :

Code:
R9600 :     300MHz Core; 200MHz RAM on R9200-PCB
R9600Pro:   375MHz Core; 275MHz RAM on R9200Pro-PCB
R9600U :    450MHz Core; 350MHz RAM on new PCB

This would have made it possible to cover most lowend-mainstream markets with one chip. The lower end card would be really cheap to produce and would give at least the same performance as the R9200Pro and R8500 but DX9 and the R9600-Ultra would be as fast as an R9500Pro.
So with one chip and 2 old PCB's (+one new) they would be able to cover the market from $99 till $229.

So why hasn't ATi done this? Or is my understanding completely wrong?

Remember what happened to the Geforce 4 ti4400? Most people were either looking for the budget version or the high end version (ti4200 and ti4600, respectively) so relatively few people actually bought it.
 
Ostsol said:
Remember what happened to the Geforce 4 ti4400? Most people were either looking for the budget version or the high end version (ti4200 and ti4600, respectively) so relatively few people actually bought it.

Well, I think it had more to do with stupid pricing. At least here in sweden the Ti4400 was just a little bit less expensive than the Ti4600. The Ti4200 on the other hand was substatialy cheaper than the Ti4400.
 
I bought a Ti4400 as it was nearly £100 cheaper than the equivalent Ti4600. And it overclocked to Ti4600 and above speeds very easily. It was a bit of a bargain really.
 
I was looking @ the OC of 9500 pro cards, and finally i don't find the OC 9600 pro really impressive :(. Sure it's over 500 MHz, but 9500 pro can do as much as 50% more on the clock rate, so....
 
Evildeus said:
I was looking @ the OC of 9500 pro cards, and finally i don't find the OC 9600 pro really impressive :(. Sure it's over 500 MHz, but 9500 pro can do as much as 50% more on the clock rate, so....

They are both better than anything nVidia makes...... ;)
 
martrox said:
Evildeus said:
I was looking @ the OC of 9500 pro cards, and finally i don't find the OC 9600 pro really impressive :(. Sure it's over 500 MHz, but 9500 pro can do as much as 50% more on the clock rate, so....

They are both better than anything nVidia makes...... ;)
Sorry but i don't think so. The card is OC by 30% on B3D or HFR, 40% by [H], and i can find plenty of GF4 Ti 4200 with 25-35% OC. So what's so impressive? Some figures:

http://www.hardware.fr/news/lire/17-04-2003/

Tonight with a bit of luck Dave will give us some 5600 U OCing.
 
Back
Top