Oblivion finally benchmarked!

[fierydemise]

Will someone show me where they're getting the data that there is higher demand for the 7900 then the X1900, people keep saying this but I haven't seen any facts proving this.

 

[Frostwake]

Originally posted by: fierydemise
Will someone show me where they're getting the data that there is higher demand for the 7900 then the X1900, people keep saying this but I haven't seen any facts proving this.

Theres no data, its all in their green tinted heads

 

[CaiNaM]

Originally posted by: Zstream
OMG just shutup, I read all of your responses and you point to IQ without talking about FPS.

It DOES matter if they used grass shaders the performance hit is around 5-6% and when a game averages 15-60 FPS that matters a TON! If you do not believe me run the FRIEKING game with them on and off! If you cannot do this then try google! If you still do not understand then please stop posting.

it matters 2000lbs? OMG just shut up with all this technospeak! could you put that into layman's terms for me please?

if the framerate is so crappy ask them why they didn't turn it off!

lol, some of you are too funny

 

[Cookie Monster]

Originally posted by: Ackmed
Typical bias response. I guess NV was not experiencing high demand back with the 6800U, and ATi was with the X800XT/PE. According to your logic, thats how it works.

As I said before, NV didnt have this problem with the 7800GT/GTX launch. They never sold out, and prices dropped, not increased due to lack of cards. They were in very high demand as well, yet none of the current problems now, were back then. Thus, your arguement is false.

When the 6 series was released along with the X series LOTS of people thought ATi had the edge. The success of 9700/9800 cards gave great reputation to ATi along with other stuff like better IQ etc. No the 6800U wasnt that succesful like the GT variant, but the whole 6 series was the first step to redemption for NV when their FX series lost them alot of market share plus reputation AND money (They had to convince many that there product wasnt afailure like last time).Back then, MANY had an ATi card unlike some (Rollo for example although he had a reason).

Notice why NV didnt release the NV47(G70) against the X850 refreshes? NV had 1 year stock piling "7800" chips. They also had Sli which wouldve made no sense to release the "refreshes". Even with the demand, they never sold out because there was the quantity to back it up.

As of now, MANY people think NV is better than ATi. Performance is the ONLY thing that matters to the average joe. Only you and I plus the rest of the hardware folks talk about OCing, IQ etc. The 7800GT along with the hard launch of 7800GTX pretty much delivered what the average joe wanted where ATi delayed its launch and had problems.

As i was saying, NV can produce twice as much per wafer than R580 thanks to the die shrink plus decrease in transistor count. There had been NO report of heat issue NOR power leakage. NV has moved to 90nm fully now low to high end. They are getting good yields, but the question is why are they out of stock already?

There are two possible explanations. One is high demand. Second is that they are having problems. (Third could be because the partners didnt buy enough chips). I doubt second because we would know it by now. Im guessing and think it is high demand. (Based on the current trend of market share/capitalisation and what NV is doing as of now)

"Typical bias response."
Typical what? bias? Think about it. Where do you think ALL those 7900GT/GTXs went? Down the toliet? Why do you think they are pretty much giving away those expensive X1800XT/X1900XTX?

 

[hans030390]

I didn't know you could bench Oblivion...but I guess they did.

Also, did it mention what drivers they were running? Were they using the optimized ones? If not, that may be why there is a difference in performance/IQ.

About Sm3, I think theres an option in the ini file that allows you to run it in a sm3 path. Not sure how that affects IQ or performance at all.

I think the article isn't completely believable because usually the 7900gtx performs very closely with the x1900xtx, and this is a bigger gap than I expected.

I guess we'll just have to wait for more benchmarks.

 

[Ackmed]

Oh ok. So its ok for NV to be sold out, and that means that nobody is buying ATi cards, but it doesnt work the other way. Just about what I expected you to say. The X800XT/PE was sold out, and the 6800U wasnt. Now that the roles have reversed, its not the same anymore. Funny that.

"Where do you think ALL those 7900GT/GTXs went? Down the toliet? Why do you think they are pretty much giving away those expensive X1800XT/X1900XTX?"

"All"? Could you give us some numbers, of these "all"? As I said before, the 7800GT/GTX launch did not have these problems. Im betting that there were far more 7800's available in the first month, than the 7900's. No facts to back this up, just an observation. And the 7800 series dropped in price in less than a month, just like the X1900 series. I guess they were "pretty much giving away" the 7800 series as well? No. Dont be so ignorant, if its good for the goose, its good for the gander, so to speak.

 

[NoDamage]

Originally posted by: hans030390
I didn't know you could bench Oblivion...but I guess they did.

Also, did it mention what drivers they were running? Were they using the optimized ones? If not, that may be why there is a difference in performance/IQ.

About Sm3, I think theres an option in the ini file that allows you to run it in a sm3 path. Not sure how that affects IQ or performance at all.

I think the article isn't completely believable because usually the 7900gtx performs very closely with the x1900xtx, and this is a bigger gap than I expected.

I guess we'll just have to wait for more benchmarks.

Yeah I think this is the real question with the benchmarks. What driver revisions they were using might make a big difference (both the Omega 6.3's and the NGO optimized 81.25's have reportedly produced performance increases), as well as the Direct3D render ahead/flip queue size settings in the drivers, plus any .ini tweaks like the bAllow30Shaders, iPreLoadSizeLimit, bUseHardDriveCache, etc.

If their goal was to find out what the maximum smooth/playable settings were, then they should have accounted for all these little tweaks, since anyone attempting to play on max settings will want to use them. And it is entirely possible that the maximum playable settings will increase significantly once the tweaks are taken into account, judging by reports from other users on various forums.

 

[5150Joker]

Originally posted by: Ackmed
Oh ok. So its ok for NV to be sold out, and that means that nobody is buying ATi cards, but it doesnt work the other way. Just about what I expected you to say. The X800XT/PE was sold out, and the 6800U wasnt. Now that the roles have reversed, its not the same anymore. Funny that.

"Where do you think ALL those 7900GT/GTXs went? Down the toliet? Why do you think they are pretty much giving away those expensive X1800XT/X1900XTX?"

"All"? Could you give us some numbers, of these "all"? As I said before, the 7800GT/GTX launch did not have these problems. Im betting that there were far more 7800's available in the first month, than the 7900's. No facts to back this up, just an observation. And the 7800 series dropped in price in less than a month, just like the X1900 series. I guess they were "pretty much giving away" the 7800 series as well? No. Dont be so ignorant, if its good for the goose, its good for the gander, so to speak.

If the 7900 cards were being made in mass quantity they'd be in stock everywhere much like the 7800 was. Seeing how their stock is so low these days it shows that supply isn't quite as good as nVidia fans woud like others to believe. Facts are the X1900 cards are in stock and priced lower than the GTX and selling very well - funny thing is people predicted exactly the opposite.

 

[5150Joker]

Originally posted by: CaiNaM



i have hq af. it makes no difference in oblivion. if it makes no visible difference, then is it in fact really doing "more"? or is that a difficult concept to grasp?

Cainam, you're just providing far too much ammo for me to prove what an imbecile you are. When I get back home tomorrow, I'll post two screenshots from Oblivion that clearly show HQ AF makes a very noticeable difference in the game.

Have you been living under a rock all these years? Why are video cards compared against one another if not for performance and IQ measurement? A radeon 9600 will give "adequate" performance and IQ in Oblivion, does that make it equal to a high end card? According to you it does.

The 7900 GTX OC nets a 41.6% performance gain for min fps by having grass shadows turned off. That is a HUGE performance gain for having a much lighter workload. The IQ differences provided by grass shadows is irrelevant because the fact is it costs a lot of performance to have it enabled. Since the XTX has grass shadows enabled in addition to angle independent AF (which DOES make a noticeable IQ difference and I'll gladly prove you wrong) it is under a much bigger workload. Remove HQ AF and grass shadows (while still retaining similar IQ according to you since HQ AF and grass shadows don't make a difference) the performance delta between the two cards would get even bigger. Finally factor in the fact that the 7900 GTX OC costs $560 vs. a stock XTX that costs $499, then the stomping becomes very evident: The XTX costs $60 less and outperforms the OC'd GTX while under a bigger workload.

I know it causes nVidia fans much aggravation to see their precious nVidia get thrashed so hard but the facts speak for themselves, no matter how much you try to obfuscate reality.


nV fans excuse/spin for the understocked and overpriced 7900 OC series poor performance in shader heavy titles in 2006: "It doesn't matter if the card is using a much lighter workload and still performs worse, at least it's playable!"

 

[Cookie Monster]

Bah I dont really care anymore. Its hard nor impossible to have a discussion with a person when a he or she is already convinced that he is right no matter what.

Here, have a cookie.

 

[Alaa]

Originally posted by: Cookie Monster
Bah I dont really care anymore. Its hard nor impossible to have a discussion with a person when a he or she is already convinced that he is right no matter what.

Here, have a cookie.

 

[TheRyuu]

The subtopic said "it wasn't even close". The difference was 3fps.
Seems close to me.
But, the ATI card IS using HQ AF so you kinda have to take that into account. Lucky for ATI their AF doesn't have the same performance hit as it does on NV hardware.

Overall, I'd say it was pretty close, with the X1900 coming out a "hair" in the lead. Eitherway, Oblivion will pretty much make any card chug along, even my 7900GT's in SLI.

Also, can someone explain why even with my 7900GT's ( @ 556/1800) that it still chugs and has noticable slowdowns? Is it because I have everything maxed out? Are softshadows a big hit in this game?

 

[5150Joker]

Originally posted by: wizboy11
The subtopic said "it wasn't even close". The difference was 3fps.
Seems close to me.
But, the ATI card IS using HQ AF so you kinda have to take that into account. Lucky for ATI their AF doesn't have the same performance hit as it does on NV hardware.

Overall, I'd say it was pretty close, with the X1900 coming out a "hair" in the lead. Eitherway, Oblivion will pretty much make any card chug along, even my 7900GT's in SLI.

Also, can someone explain why even with my 7900GT's ( @ 556/1800) that it still chugs and has noticable slowdowns? Is it because I have everything maxed out? Are softshadows a big hit in this game?

Did you read any of this thread? If not see my post above. The GTX was:

1. OC'd
2. Not using grass shadows which netted a 41.6% increase in min fps (12-->17 fps)
3. Not using angle independent AF (not possible)
4. Costs $60 more than a stock XTX.


That's a major thrashing for a card that costs more, has bad availability, is forced to use lower settings and STILL loses.

 

[TheRyuu]

Originally posted by: 5150Joker

Originally posted by: wizboy11
The subtopic said "it wasn't even close". The difference was 3fps.
Seems close to me.
But, the ATI card IS using HQ AF so you kinda have to take that into account. Lucky for ATI their AF doesn't have the same performance hit as it does on NV hardware.

Overall, I'd say it was pretty close, with the X1900 coming out a "hair" in the lead. Eitherway, Oblivion will pretty much make any card chug along, even my 7900GT's in SLI.

Also, can someone explain why even with my 7900GT's ( @ 556/1800) that it still chugs and has noticable slowdowns? Is it because I have everything maxed out? Are softshadows a big hit in this game?

Did you read any of this thread? If not see my post above. The GTX was:

1. OC'd
2. Not using grass shadows which netted a 41.6% increase in min fps (12-->17 fps)
3. Not using angle independent AF (not possible)
4. Costs $60 more than a stock XTX.


That's a major thrashing for a card that costs more, has bad availability, is forced to use lower settings and STILL loses.

*reads article*
...

(my dying argument)
1. ...you win
2. You don't see the shadows?
3. ...not much difference to notice
4.no argument.

Fine, you win. The X1900 is better for Oblivion. Now, is it because of the 48 pixel shaders? or other ATI hardware? Now I just want to know why?

But I still like my 7900GT's.

 

[Dainas]

The jist of it is that the X1900 architechture will always win where pixel performance is more important that texturing. ATI ineptness at Opengl is the only thing that keeps nvidia ahead at the doom 3 engine.

Nvidia still needs to straighten their act out on AA/anis. While I hate looking at the occasional blah texture, oblivion is incredible. 1600x1200 16anis+HDR, all sliders and options on and the only thing that slows it down below 30fps is heavy AI outside.

 

[Yreka]

Originally posted by: mazeroth

it's pretty pathetic when ppl have to grasp at every little straw to try and claim an advantage. while i'd have to agree the xtx has an edge (also cost more, btw), the reality is they offer the same gameplay and same iq (tho you have to "suffer" with a little lighter grass texture if you have a GTX).

Actually, the 7900 GTX is CONSIDERABLY more expensive than the X1900XT. Yes, I know the review used the X1900XTX, but I've never seen an X1900XT that can't reach XTX speeds, ever. The X1900XT can be had for $100 less than a 7900 GTX, so the comparison is way more in favor of the ATI. No, I'm not a fan of either as I have a 7800GT and an X1900XT.

Mine wouldnt reach XTX with the stock HSF. It has the 1.1 memory too, I was limited by temp.

 

[LittleNemoNES]

Originally posted by: Yreka

Originally posted by: mazeroth

it's pretty pathetic when ppl have to grasp at every little straw to try and claim an advantage. while i'd have to agree the xtx has an edge (also cost more, btw), the reality is they offer the same gameplay and same iq (tho you have to "suffer" with a little lighter grass texture if you have a GTX).

Actually, the 7900 GTX is CONSIDERABLY more expensive than the X1900XT. Yes, I know the review used the X1900XTX, but I've never seen an X1900XT that can't reach XTX speeds, ever. The X1900XT can be had for $100 less than a 7900 GTX, so the comparison is way more in favor of the ATI. No, I'm not a fan of either as I have a 7800GT and an X1900XT.

Mine wouldnt reach XTX with the stock HSF. It has the 1.1 memory too, I was limited by temp.

Then you're card was the exception. I even put the XTX bios on mine. Bought my x1900xt for $450. Quite a deal.

 

[Clauzii]

Originally posted by: CaiNaM

Originally posted by: 5150Joker
The X1900 XTX had grass shadows enabled in addition to HQ AF (something nVidia can't even reproduce) and it STILL beat out the 7900 GTX in Oblivion, that's pretty significant. If they did an apples to apples comparison and turned off grass shadows for the XTX and used standard AF it would've stomped the 7900 GTX; the GTX went from a minimum of 12 fps to 17 fps by shutting off grass shadows.

a feature with no benefit is hardly a feature:

The difference between the image quality on the Radeon X1900XTX and GeForce 7900 GTX wasn't noticeable, even with Grass Shadows turned off - all that seemed to do was darken the grass texture a little.

it's pretty pathetic when ppl have to grasp at every little straw to try and claim an advantage. while i'd have to agree the xtx has an edge (also cost more, btw), the reality is they offer the same gameplay and same iq (tho you have to "suffer" with a little lighter grass texture if you have a GTX).

And given the fact that the X1900XTX runs at a higher resolution makes You´r statement(s) seem like bable to me :roll:

 

[coldpower27]

Originally posted by: Clauzii

Originally posted by: CaiNaM

Originally posted by: 5150Joker
The X1900 XTX had grass shadows enabled in addition to HQ AF (something nVidia can't even reproduce) and it STILL beat out the 7900 GTX in Oblivion, that's pretty significant. If they did an apples to apples comparison and turned off grass shadows for the XTX and used standard AF it would've stomped the 7900 GTX; the GTX went from a minimum of 12 fps to 17 fps by shutting off grass shadows.

a feature with no benefit is hardly a feature:

The difference between the image quality on the Radeon X1900XTX and GeForce 7900 GTX wasn't noticeable, even with Grass Shadows turned off - all that seemed to do was darken the grass texture a little.

it's pretty pathetic when ppl have to grasp at every little straw to try and claim an advantage. while i'd have to agree the xtx has an edge (also cost more, btw), the reality is they offer the same gameplay and same iq (tho you have to "suffer" with a little lighter grass texture if you have a GTX).

And given the fact that the X1900XTX runs at a higher resolution makes You´r statement(s) seem like bable to me :roll:

X1900XTX runs the same resolution as 7900GTX

 

[akugami]

It's funny how people are blinded by colored shades. The 7800 was THE best card out when it came out and it had much much better availability. Seriously, there was no question it was the best game in town. Anyone who says otherwise is an idiot. That means that it should have sold at a better rate than the 7900 which is a very slight 2nd fiddle to the X1900 according to most, or a very slight first according to some. The difference is not great enough that you can really go wrong choosing either card. So nVidia with an "inferior" card (7900) it's selling at a better rate than when it had the undisputed absolute performance leader (7800) for about 6-8 months? And I'm saying inferior as in it's not the best card compared to the competition, not saying the 7900 is a bad card at all. In fact, as I say, the X1900's lead is a very slight one.

Now, it is possible the 7900 is a better selling and more popular card than the 7800 but I really doubt it. Most people follow the performance and most web sites and magazines have the X1900 as the leader. But if as you say, most people will buy the performance leader wouldn't that mean that the X1900 has a greater share of sales compared to the 7900 than the X800 series compared to the 7800's? The 7900's also have moved to a 90nm process which would produce more gpu cores per wafer as compared to the 7800's on a 110nm process. That's assuming similar yields of course and unless there is a manufactoring problem, the yields should be about the same. That would mean that nVidia should be able to produce more 7900's than 7800's. That should mean the supply of 7800's should have been tighter than the 7900 series and not more available. The 7900's are also on a new (for nVidia) 90nm process as well and moving to 90nm has given everyone problems from most notably some of the biggest tech corps in IBM, Intel and AMD. I'd say this points to a supply problem rather than extremely high demand for the 7900's.

Either way, it's all speculation unless someone gets hard data and quite frankly, the video card market is one where data is extremely hard to come by. We can get numbers like overall gross sales for nVidia and ATI and we get stuff like how many discrete and how many integrated chipsets each company sells but we don't get stuff like how many high, mid and low end cards they sell. Which is what we need to determine who is the better seller and in what category. So for some who swear up, down, left right that nVidia is selling a lot more video cards than ATI in the highe end you can't prove it so it's not fact. Merely speculation. I can just as easily say that ATI is out selling nVidia 2 to 1 in the high end. You can't prove me wrong either because the data simply is not there. Or at least if the data is there, it's not available to the general public, at least not where I can find it. If the data is out there I'd love to see it. Not because of any fanboy attachement to either company. I'm just curious as to who sells what and how much.

 

[RobertR1]

The puzzling part in the article is the use of grass shadows which they claim are performance impacting with minimal visual gains as clearly inidcated by their conclusion.

Then why even turn on Grass Shadowing? Who's to say that the x1900xtx wouldn't run those FPS, excluding grass shadows at even 1600x1200, seeing the 40%+ hit from grass shadows on nvidia hardware.

 

[CaiNaM]

Originally posted by: Clauzii

Originally posted by: CaiNaM

Originally posted by: 5150Joker
The X1900 XTX had grass shadows enabled in addition to HQ AF (something nVidia can't even reproduce) and it STILL beat out the 7900 GTX in Oblivion, that's pretty significant. If they did an apples to apples comparison and turned off grass shadows for the XTX and used standard AF it would've stomped the 7900 GTX; the GTX went from a minimum of 12 fps to 17 fps by shutting off grass shadows.

a feature with no benefit is hardly a feature:

The difference between the image quality on the Radeon X1900XTX and GeForce 7900 GTX wasn't noticeable, even with Grass Shadows turned off - all that seemed to do was darken the grass texture a little.

it's pretty pathetic when ppl have to grasp at every little straw to try and claim an advantage. while i'd have to agree the xtx has an edge (also cost more, btw), the reality is they offer the same gameplay and same iq (tho you have to "suffer" with a little lighter grass texture if you have a GTX).

And given the fact that the X1900XTX runs at a higher resolution makes You´r statement(s) seem like bable to me :roll:

huh? the article they are running same resolution.

 

[Doom Machine]

nvidia does wonders with their drivers...perhaps a patch from bethseda and drivers from nvidia will no doubt get it much smoother. bethseda said they worked close with nvidia on drivers..whatever thats suppose to mean

 

[FalllenAngell]

Originally posted by: zzzvideocardzzz

Originally posted by: nib95
Well, I'm running mine at nearly everything maxed 1920 x 1200 and my average frame rate is 60fps.
60-160fps in dungeons, 40-100fps in cities, and 40-60fps outdoors.

SLI FTW!

and exactly how many ppl can afford 600 dollars? and the extra money on SLI mobo and sli psu? and we all know how SLi setups does badly in mininum framerate department. Just buy a x1900xt for around 400 dollars now and oc it to the max FTW

Millions of people bought SLI motherboards when they upgraded to PCIE, because it just made sense to do so if you are an enthusiast. I seriously doubt most people with $400-$600 video cards bought single slot motherboards.

As far as a PSU goes, with the X1900 XTX requiring a whopping 120Watts at full load, and the 7800GTXat 80 watts, I don't think anyone is saying "Yay! I get to keep my 350W PSU from 1998 to run my X1900XT!".

Not everyone wants to "OC cards to the max" . The cheapest X1900XT on newegg is $434., which to me, is a fair amount of money to risk "OCing to the max".

As reported at DailyTech:
ATI claims thermal damage is strictly not covered by RMA and that blown internal fuses in the GPU will notify RMA teams if the GPU has been stressed too hard.

If other brands can do the same (check the blown internal fuses) I personally don't want to be the guy they're telling,"Too bad, so sad. The blown internal fuses indicate you've "OCd your card to the max" and this is now considered a fraudulent RMA."

 

[the Chase]

Originally posted by: Cookie Monster

Originally posted by: Ackmed
Typical bias response. I guess NV was not experiencing high demand back with the 6800U, and ATi was with the X800XT/PE. According to your logic, thats how it works.

As I said before, NV didnt have this problem with the 7800GT/GTX launch. They never sold out, and prices dropped, not increased due to lack of cards. They were in very high demand as well, yet none of the current problems now, were back then. Thus, your arguement is false.

When the 6 series was released along with the X series LOTS of people thought ATi had the edge. The success of 9700/9800 cards gave great reputation to ATi along with other stuff like better IQ etc. No the 6800U wasnt that succesful like the GT variant, but the whole 6 series was the first step to redemption for NV when their FX series lost them alot of market share plus reputation AND money (They had to convince many that there product wasnt afailure like last time).Back then, MANY had an ATi card unlike some (Rollo for example although he had a reason).

Notice why NV didnt release the NV47(G70) against the X850 refreshes? NV had 1 year stock piling "7800" chips. They also had Sli which wouldve made no sense to release the "refreshes". Even with the demand, they never sold out because there was the quantity to back it up.

As of now, MANY people think NV is better than ATi. Performance is the ONLY thing that matters to the average joe. Only you and I plus the rest of the hardware folks talk about OCing, IQ etc. The 7800GT along with the hard launch of 7800GTX pretty much delivered what the average joe wanted where ATi delayed its launch and had problems.

As i was saying, NV can produce twice as much per wafer than R580 thanks to the die shrink plus decrease in transistor count. There had been NO report of heat issue NOR power leakage. NV has moved to 90nm fully now low to high end. They are getting good yields, but the question is why are they out of stock already?

There are two possible explanations. One is high demand. Second is that they are having problems. (Third could be because the partners didnt buy enough chips). I doubt second because we would know it by now. Im guessing and think it is high demand. (Based on the current trend of market share/capitalisation and what NV is doing as of now)

"Typical bias response."
Typical what? bias? Think about it. Where do you think ALL those 7900GT/GTXs went? Down the toliet? Why do you think they are pretty much giving away those expensive X1800XT/X1900XTX?

You hit the major reason.(My beleif). #3. Can't find the Inquirer post on it now but they wrote how the Nvidia board partners weren't that impressed with the 7900 chip and so limited their orders. They are probably now kicking themselves. I think the demand was there but they haven't been able to replenish stock at all really for what, 2 weeks or more now. Something else other than demand is playing a role here.