When comparing GF FX to 9700PRO, why dont benchmarkers do this....

[BoomAM]

Yeh, whatever, but Gainward doesnt make every single GF FX card out does it?
Have you seen the card shots of the other manufacturers? They`re just using the vacuum cleaner.

Becides this is straying off the subject.

 

[BFG10K]

Do you see where im going with this people.

Yeah I do but it won't achieve anything since nobody buys the cards in the configurations that you're suggesting to use in the tests.

It's no different to taking a 2 GHz P4 and clocking it down to 1.667 GHz to match a 2000 XP. Sure you can do that but it's pretty pointless. A much more reasonable approach would be to test 1.6 , 1.8 and 2.0 GHz variants to see how they stack up against the 2000 XP, variants that are actually available to the public to buy. What you're suggesting is it make up an imaginary video card(s) that nobody can ever get their hands on.

It might be good for a theoretical test but not much more than that.

We could also see which core was more efficient as well.

Not really since the efficiency of the core could include the ability to scale to high clock speeds.

 

[KingofFah]

I would like to see this benchmark, it show somewhat relevant results of the future.
If we slowed the R300 down to match the bandwidth, the ratios would probably be the same as comparing an r300 to an nv30 with a 256bit bus.
Is this benchmark really worth it?
No, it is just to satisfy a curiosity about the cores, but I think we know what the answer would be.
nVidia was real stupid this time, and they are paying for it right now (literally, hehe). Had they implemented a 256bit bus, we would be seeing some amazing performance.

 

[kylebisme]

the thing is that while we do know which way it comes down, we don't know how much to one side; that is what would be intersting to see and why such a test would be good to do.

 

[dxkj]

Umm guys..... The Radeon would win?


Why are you even asking this question Boom? You KNOW that the radeon would win, hands down, that is why the Geforce is clocked so much faster..



This isnt a question of which core or design is more efficient, we all KNOW that the ATI core is more efficient in terms of speed....




So yes this is a pointless thread IMO where you are simply trying to get everyone to say All Hail ATI, they are more efficient...


And yes I agree I dont think ANYONE would disagree that the R300 core is > FX in efficiency per Clock...

 

[kylebisme]

did you even read what o wrote dxkj?


*feels ignored*

 

[UncleWai]

It makes no sense. The two companies take different approaches to advance their products. Nvidia is going on the route of Intel which is to pump the highest clock possible to achieve performance. On the other hand, ATI is focusing on the architecture of the GPU so it can do more things at lower clock.
Your comparison is abandoning clockspeed, which is the key component for Nvidia to achieve speed.
It can only prove which GPU is better built architecture wise, which is pretty obvious.

You can switch the comparison around and have both GPU running at the default clock of FX 5800 Ultra.
Oh nevermind, 9700 pro can't clock that high.

 

[kylebisme]

because it is not .13nm and sure it doesn't have any value when comparing the two products themselves. the point is in seeing the difference in how the architecture itself preforms when those differences are negated. if you all do not find the subject interesting that is fine, but telling us it is stupid to wonder is crossing the line imho.

 

[BoomAM]

Originally posted by: TheSnowman
because it is not .13nm and sure it doesn't have any value when comparing the two products themselves. the point is in seeing the difference in how the architecture itself preforms when those differences are negated. if you all do not find the subject interesting that is fine, but telling us it is stupid to wonder is crossing the line imho.

Exactely.
Just because you lot see it as a pointless exercise, is no reason do deem it completely useless, and slate people who would find it interesting.
Upto now, TheSnowmans&Sunner is the only other people who seem to appriciate what im suggesting.

 

[HappyNic]

When you change the GPU/VPU mhz or their memory bandwidth you are changing how they were design to work. By doing so your answer would be meaningless because they weren't desgin to work that way. What you want is not real comparsion but Paper Specs. But every know its' all about REAL test, REAL RESULTs.

I'll give a example: American Boxing vs Tai boxing. If they fight by their own style,, most likey TAI would win. But if don't allow knee, kick, elbow, etc,,, to make it same as normal Boxing then of course American Boxing would mostly likey win., but how about what they have beening Eating, or the girls they sleep with, or the Death of their FAmily, ETc, etc.. ALL of them have to be the same by your meaning,, it's imposiable to compare.


Just like for chips, .13 Micron vs .15 micron, should we inlarge or decease the size on both, and the millions of Transitors on the chip, should we remove or add some on the other chip to make them both even? How about power should we only give the same power to run the chips? what about the cooling solution, both have to use the excat same heatsink and fan running at the same speed. Both Chip have to be the excat same size made but excat same people. in the Exact area,. ETC, ETC

Once agian, What you are asking is really stupid because the answer would mean nothing if you can't get everything excatly the same. If it's ok for .13 vs .15 micron why can't it be ok for 325 vs 500mhz. If it's ok to have ddr2 vs ddr1, why can't they have 128 vs 256 bit.

ARe you getting what I'm saying..... pork chop is different from Steak.

 

[kylebisme]

i get what you are saying HappyNic, i just dont find it convenceing. your arguments seem extremely sophistical to me, beyond being blatantly insulting.

 

[HappyNic]

Originally posted by: TheSnowman
i get what you are saying HappyNic, i just dont find it convenceing. your arguments seem extremely sophistical to me, beyond being blatantly insulting.

I' just telling BoomAM straight to the point and if it seem insulting then I guess the truth does hurts sometimes. It wan't ment to be insulting, it was mean to be "don't set werid restrictions for something that shouldn't be compared in that way?"

 

[kylebisme]

well i really don't see where you get off saying you pretending to be some authority on what "should" or "shouldn't be." as a matter of fact i will go so far as to say you are completely out of line and arrogant far beyond reason to even posture yourself in such a conceded manor. how does that truth go over on you HappyNic?

 

[beerbong]

i remember seeing a comparison of the 9700vsParhelia ..both same clockrate. The parhelia had better scores in many benchmarks. TO bad the Parhelia clockrate isnt higher.
beer

 

[Bacinator]

Originally posted by: BoomAM
What is is with ppl here lately, everyone seems to be throwing insults like they`re going out of fashion.

What alot of you lot seem to be missing, is that its just for research and stuff.
Like, would the 9700pro be better than a GF FX, given the same bandwidth limitations?
Its all very well comparing my suggestion to apples & engines, but you are way off my inital point.

Rortary and 4 stoke engines can use the same gastank, exhaust, etc... My point is still that some of the major functions are different, and potentially incomperable under the same scale/tests. Hench the varience/availbilities of multiple benchmarks and testing methods.

 

[kylebisme]

but it is still nice to run them on the same octane fuel and the same quality of oil, ignition, sparkplugs and the like in order to comparatively gauge the effectiveness of the engines themselves. do you not see the logic in that Bacinator?

 

[BoomAM]

When everyones finished comparing the GFX cards to a engine, or fruit, or something else. Its getting daft now. We dont need no more comparisons.

 

[BoomAM]

Originally posted by: TheSnowman
but it is still nice to run them on the same octane fuel and the same quality of oil, ignition, sparkplugs and the like in order to comparatively gauge the effectiveness of the engines themselves. do you not see the logic in that Bacinator?

LOL.
I like this guy. Him funny.

 

[kylebisme]

well thank you BoomAM, you seem like a quality fellow yourself.

 

[BoomAM]

Why thank you yourself, dear sir!!!
Ive gotta get that seen to. lol.

 

[FuManStan]

I don't see why people are shrugging it off as completely useless. It's merely a test that could be of interest to some people. There HAVE been reviews that have compared the P4 to the Athlon on a clock for clock basis. In fact, many reviews reiterate the fact that the Athlon is faster when comparing clock speeds. Does that fact matter in the end? Not at all, but its mentioned regardless. I see no reason why the same can't be applied to graphics cards.

 

[kylebisme]

FuManStan, like i said above, i belive the reasoning behind such actions is put quite plainly in the lyrics to this song found here.

 

[Lonyo]

Wasn't it said that the Gainward solution was 7dB QUIETER than the normal cooler?

You say we should use same core/mem clock, how about making them same noise level as well and see how they perform?
Or how about mounting on the same PCB?
How about seeing which can move itself over a stretch of water quickest?

What I'm saying is there's no point in using the cards at equal speeds to compare, because they're not equal in that way.
You're trying to hinder them memory bandwidth of the 9700PRO to see how it does with less, when the simple fact is it DOESN'T HAVE LESS, so you're making it unfair, you're taking away things that make it what it is.
I don't know how to explain what I mean properly, but what you suggest is just wrong.

You're saying compare two cards by making their speeds the same, when the cards you'd go out and buy aren't the same speed, what's the point?
As some one said, it's like clocking an Athlon with higher FSB, and lowering the speed of a P4 so they match, when that's not what happens in real life, so it's totally pointless in terms of what you really see.

 

[kylebisme]

...

I scramble to reach higher ground,
some order and sanity,
or something to comfort me.

So I take what is mine, and hold what is mine,
suffocate what is mine, and bury what's mine.


...

 

[BoomAM]

Lonyo,
The noise level of the cooler will not affect the cards performance.
Mounting a 9700 on a GF FX PCB would make it a fast 9500.
Mounting a GF FX on a 9700 PCB, would give it a 256bit memory bus. Which would be interesting. But i dont know of any review site that has the resources to do this. Let alone the permission/s off ATI/nVidia.
Computer components cant swim.