id Software's Official DOOM3 Benchmarks

[Tremek]

I think you're lost. The Democratic Underground is over thataway.

<<<---------lll

Mistaking treehuggers for Democrats earns you a metaphorical whallop in the jewels. Don't be ignorant! Talk more about Doom3 instead of making stupid political non-commentary.

 

[Chris_B]

after reading some of your posts on this thread they seem pretty retarded like this one. even if the 60fps is capped if i could get 80fps on D3, i'd welcome it. the more the better since i could do more res and lower fps drops and yes even with 5fps more you would notice the difference.

any hardcore fps'er would know that.

Yawn dont bore me child.

Retarded meaning dont trust leaks for reliable benchmarking, yep very retarded advice. Lets all go by leaks, on the stalker leak i get 500fps, that must mean ill get 100fps when it comes out gg.

 

[PucSpifo]

a couple of people like who cares about 20 fps etc. etc.? while i cannot see the diffrence i can feel the diffrence between 80 and 60 fps, i play massive amounts of dod on my 9800 pro and when im running other programs as well everyonce in a while it will dip down to 60 and i catch it every time looking down at netgraph going wtf, so yes, 20 fps, or even 5 fps for that matter, matters

Of course 90% of the world can't recognize motions above 60-66 FPS, and many have a very hard time with anything above 30 FPS. When you watch TV how many FPS are you seeing? 25 (PAL) to 30 (NTSC, actually 29.4). So guess what, above 60 FPS won't matter, and the real sweet spot for most is 30 FPS.

 

[rancor]

Of course the pipelines have the necessary transistors to perform FP32 on the NV3x line. It's silly to say they don't. The number of pixel and vertex shaders don't rely on the width of the FP data. That's like saying an Athlon 64 needs twice as many executions units to run 64-bit software. LOL

I think the bandwidth comment was just to rub in the fact, at the time, that NV was being boneheaded about going to a 256-bit memory bus. I'll bet that when the comment was made the speaker had no idea the real problem was pixel shader 2.0 performance. The NV30 was actually pretty good in DX8.x and DX7 performance, but lagged the R300 badly in PS2.0 performance.

What are you talking about? Sorry for being rude, just that I'm foward, but That article was after the 6800 release and the x800 release.

Transistor counts have to do with the shaders, the graphics pipelines use pixel shaders and vertex shaders regardless if there is a programmable shader program for them. if they weren't used there would be nothing drawn to the screen. Those transistors are the shaders. What do you think constitues the increase in number of transistors for the Dx9c's new shader models with more complexitiy? If the transitors had nothing to do with shaders then you wouldn't need those extra 60 million transistors for dx9c.

Even GF 1's had shaders, just werent' programmable.

Sorry forgot to mention all textures are 32 bit to begin with anyways. So 24 fp is not really color data in a texture is the amount of caculations being done on that 32 bit texture.

So the transistors really matter big time.

 

[Tremek]

Of course 90% of the world can't recognize motions above 60-66 FPS, and many have a very hard time with anything above 30 FPS. When you watch TV how many FPS are you seeing? 25 (PAL) to 30 (NTSC, actually 29.4). So guess what, above 60 FPS won't matter, and the real sweet spot for most is 30 FPS.

That's total misinformation. The human eye perceives the typical cinema film motion as being fluid at about 18fps, because of its blurring. But that's not where it ends. Consider this: the human eye doesn't see in frames, it sees in a continual flow of light and information. The Air Force did tests where pilots could percieve and identify the sillouette of a plane that was flashed in front of them for only 1/220th of a second. The basic indication here (and total misinformation propigated by the concept that once you hit that magic 30-fps, the eye can't see anything else) is that the human eye can potentially differentiate between something as high as 500 fps - as in, no perception of difference from frame to frame due to flicker. Would it be considerably different than, say, 120 fps? Probably not. But don't try to continue the lie that 30 fps is not only enough, but that any more is just excess. It isn't.

 

[pxc]

What are you talking about? Sorry for being rude, just that I'm foward, but That article was after the 6800 release and the x800 release.
...
So the transistors really matter big time.

You are not relaying the point you keep referring to. Please give a link to the original quote.

I cut out the meat of your post to save you potential embarassment.

 

[BoogerBomb]

Good god I went to bed on page 14 and here it is on page 22 with the same arguments still ensuing. Why dont we just wait til [H] releases their benchmarks.

 

[rancor]

You are not relaying the point you keep referring to. Please give a link to the original quote.

I cut out the meat of your post to save you potential embarassment.

Ok here ya go

Here is the link and

http://techreport.com/etc/2004q2/nalasco/index.x?pg=2

and here is the quote from that link

One thing to consider when you're looking at these things is, you know, it might not sound like a huge difference between 24 bits or 32 bits, but what you're basically talking about is 33% more hardware that's required. You need 33% more transistors. Your data paths have to be 33% wider. Your registers have to be 33% wider. To put that into perspective, that's kind of equivalent to adding another four pipelines onto an X800 type of chip to support this additional precision. And then we have to make the choice of, "What kind of visual benefit is that going to provide? What kind of programming benefit is that going to provide?" and weigh that against improving performance.

I'm stating the facts there is no embaressment here .

 

[primea]

This was a post about Doom3 benchmark, how it will allow older video cards to run it ... and the second post started the Ati vs. Nvidia banter ... did nobody read the article ...

 

[rancor]

This was a post about Doom3 benchmark, how it will allow older video cards to run it ... and the second post started the Ati vs. Nvidia banter ... did nobody read the article ...

Well I was getting at Ogl doesn't support 24 fp natively thats why its not a driver issue on ATi's side was getting at. This is speculation, but just an interesting tid bit

 

[deadsoulz]

This Benchmark isn't what made me change my mind, I went out and Bought a 6800 GT a couple of days ago, because of the great price for the value it gives, unlike ATI , Nvidia gives us all 16 pipelines on this GT card and this thing overclocks with no effort to the Ultra Card Speeds. Not to metion the DirectX 9.0c support , and This new benchmark just puts the icing on the cake. I have never been a fanboy, I still love my 9800 Pro AIW thats in my Wifes machine, its a great card, I always buy whats best for the buck, being a loyalist can bite you in the ass. And if you did go out and buy a X800 Pro shame on you, if you bought a X800 XT then thats still a great card with all 16 pipelines, I imagine it will shine in opengl in the future.

--DeaDSOuLz

 

[Tremek]

More FPS information:

The overwhelming solution to a more realistic game play, or computer video has been to push the human eye past the misconception of only being able to perceive 30 FPS. Pushing the Human Eye past 30 FPS to 60 FPS and even 120 FPS is possible, ask the video card manufacturers, an eye doctor, or a Physiologist. We as humans CAN and DO see more than 60 frames a second.

Further on in that article:

This is where this article gets even longer, but read on, please. I will explain to you how the Human Eye can perceive much past the mis conception of 30 FPS and well past 60 FPS, even surpassing 200 FPS.

Read it if you're even thinking about contesting this.

http://amo.net/NT/02-21-01FPS.html

 

[Tremek]

... And if you did go out and buy a X800 Pro shame on you, ...

Shame on [you]? You're telling people they're somehow WRONG, that they've somehow made a mistake big enough that not only do you have to comment on it, but that it affects them and others as well? Get real. I hope you just don't know what you're writing.

 

[rancor]

More FPS information:



Further on in that article:

Read it if you're even thinking about contesting this.

http://amo.net/NT/02-21-01FPS.html

hmm the x800 pro is under 30 fps :/ and yeah that will make a difference.

 

[deadsoulz]

Shame on [you]? You're telling people they're somehow WRONG, that they've somehow made a mistake big enough that not only do you have to comment on it, but that it affects them and others as well? Get real. I hope you just don't know what you're writing.

Its really more my way of saying shame on ATi why didn't they leave all 16 pipelines enabled and just lower the clock speed, I think it was a bad move.

 

[BoogerBomb]

Its really more my way of saying shame on ATi why didn't they leave all 16 pipelines enabled and just lower the clock speed, I think it was a bad move.

Coulda, woulda, shoulda. Hindsite is always 20/20. I forsee X800Pro V2 coming soon.

 

[eXzite]

Its really more my way of saying shame on ATi why didn't they leave all 16 pipelines enabled and just lower the clock speed, I think it was a bad move.

They didn't leave all 16 pipes enabled because of yield. You see, when they fab these chips, all of them are potentially 16-pipeline XT's, however, since the silicon fab process isn't perfect, then you get transistor errors occasionally. Since there's hundres of millions of transistors on the die, there's a good chance at least a few will be dead.

Now, what do they do with this chip? They don't throw it away, they just disable the batch of 4 pipelines that have the error, and sell it as an X800 Pro instead. All semi-conductor companies do this. Each chip is physically different, and can therefore clock to different speeds, use different amounts of power, produce different amounts of heat, and have slightly different capability. The X800 XT PE's are just closer to what's called a 'golden sample', or chips that are absolutely perfect. ATi (and nVidia) just sell each chip at the highest level of performance they can get out of it. Coincidentally, this is why you can overclock your card, because the silicon might not stand up to the additional 50MHz of clock speed that would make it a Ultra or PE, but it can still take a higher clock that a normal GT or Pro.

nVidia had a better fab run than ATi, thefore, nVidia gets to have more chips with 16 pipes.

Coulda, woulda, shoulda. Hindsite is always 20/20. I forsee X800Pro V2 coming soon.

Actually that won't happen, because each chip only goes to fab once. A chip fab is the single most expensive thing a company can undertake, so ATi isn't going to go off running another r420 spin anytime soon.

 

[Ardrid]

They didn't leave all 16 pipes enabled because of yield. You see, when they fab these chips, all of them are potentially 16-pipeline XT's, however, since the silicon fab process isn't perfect, then you get transistor errors occasionally. Since there's hundres of millions of transistors on the die, there's a good chance at least a few will be dead.

Now, what do they do with this chip? They don't throw it away, they just disable the batch of 4 pipelines that have the error, and sell it as an X800 Pro instead. All semi-conductor companies do this. Each chip is physically different, and can therefore clock to different speeds, use different amounts of power, produce different amounts of heat, and have slightly different capability. The X800 XT PE's are just closer to what's called a 'golden sample', or chips that are absolutely perfect. ATi (and nVidia) just sell each chip at the highest level of performance they can get out of it. Coincidentally, this is why you can overclock your card, because the silicon might not stand up to the additional 50MHz of clock speed that would make it a Ultra or PE, but it can still take a higher clock that a normal GT or Pro.

nVidia had a better fab run than ATi, thefore, nVidia gets to have more chips with 16 pipes.



Actually that won't happen, because each chip only goes to fab once. A chip fab is the single most expensive thing a company can undertake, so ATi isn't going to go off running another r420 spin anytime soon.

You haven't heard the rumors about the X800 GT then

 

[KakimotoR]

My System spec should do fine with DOOM3

 

[eXzite]

You haven't heard the rumors about the X800 GT then

Acually, ATi was just like.. dammit, this batch of chips over here don't have any transistor errors in them, so we have 16 pipes, but the silicon impurities are creating a critical path that forces us to run below XT core speed. Wait.. I know, we'll sell them as a lower-clocked 16-pipe version and call it a X800 GT. Same chip guys... different marketing spin.

 

[HighHardOne]

The problem is there's absolutely no basis for the argument that 'ATI might/will perform better in other games'. Look at the current DX9 performance. NVIDIA and ATI are neck and neck and NVIDIA keeps improving with every driver release. The argument simply doesn't hold any water.

Ahh, I think you might want to read what I wrote again and keep in mind that I am not advocating anyone's position besides the consumer. I'm not interested in who has the better card for Doom3. The bottom line is the game runs well on a multitude of cards. My insight here is nVidia developing specifically for one game and basing the success/failure of the generation of cards around a single title. I'm sure shareholders wouldn't be very excited about it. I'd hate to see nVidia make anymore missteps that might increase mind/marketshare for ATi, because what we need is competition between the two. The Radeon 9700 Pro was the best thing to happen to the industry since the original Voodoo chipset. Not because it was faster, but because it forced nVidia to pay attention to their original market segment.

 

[Ardrid]

Acually, ATi was just like.. dammit, this batch of chips over here don't have any transistor errors in them, so we have 16 pipes, but the silicon impurities are creating a critical path that forces us to run below XT core speed. Wait.. I know, we'll sell them as a lower-clocked 16-pipe version and call it a X800 GT. Same chip guys... different marketing spin.

Oh, I know it's not physically a different chip. But it does mean that there'll be less fab space for other chips. It also means that ATI's going to have some strange pricing, in addition to some major cannibalization.

 

[Ardrid]

Ahh, I think you might want to read what I wrote again and keep in mind that I am not advocating anyone's position besides the consumer. I'm not interested in who has the better card for Doom3. The bottom line is the game runs well on a multitude of cards. My insight here is nVidia developing specifically for one game and basing the success/failure of the generation of cards around a single title. I'm sure shareholders wouldn't be very excited about it. I'd hate to see nVidia make anymore missteps that might increase mind/marketshare for ATi, because what we need is competition between the two. The Radeon 9700 Pro was the best thing to happen to the industry since the original Voodoo chipset. Not because it was faster, but because it forced nVidia to pay attention to their original market segment.

I wasn't specifically addressing your post, but rather the hypothetical you stated, which is what a lot of ppl are using to validate their views. That hypothetical is what doesn't hold water given what's going on now and what's likely to continue over the next 6 months.

 

[Kared]

Frankly I'm not suprised by these numbers.....and I still bought an X800Pro.

Why? Because while Doom3 is favoring the nvidia cards, Half Life 2 will be favoring the ATI cards. This is mainly due to the deals that Valve and ID signed with these respective game companies.

Since I'll be playing Half Life 2 longer than Doom3 (mainly becuse of its multiplayer and the various Mods that will be coming out for it) I still felt that ATI was the way to go.

Kared

 

[pxc]

Ok here ya go

Here is the link and

http://techreport.com/etc/2004q2/nalasco/index.x?pg=2

and here is the quote from that link

One thing to consider when you're looking at these things is, you know, it might not sound like a huge difference between 24 bits or 32 bits, but what you're basically talking about is 33% more hardware that's required. You need 33% more transistors. Your data paths have to be 33% wider. Your registers have to be 33% wider. To put that into perspective, that's kind of equivalent to adding another four pipelines onto an X800 type of chip to support this additional precision. And then we have to make the choice of, "What kind of visual benefit is that going to provide? What kind of programming benefit is that going to provide?" and weigh that against improving performance.

I'm stating the facts there is no embaressment here .

None of that contradicts what I wrote above. 1) The NV40 does have the transistors for FP32 throughout the pipeline (ATI decided not to for the reason given above, but nvidia did), 2) it doesn't require extra shader units (not even suggested in your quote) and 3) it said nothing about bandwidth (which isn't an issue as you suggested he said)

I'll remind you of your posts:
"Originally Posted by rancor
Well with 32 bit there is a incease of bandwidth that is needed, I think it was ATi's CEO that said this thats why the Fx line got screwed, they had everything but the bandwidth necessary to do it. Thats why they haven't done it yet."
[he didn't say that. the words "bandwidth" or "FX" don't appear in the entire article, and in no context does "competitor" or "competition" (the only 2 words used to describe the other company in the article) mention anything like you claimed]

"Originally Posted by rancor
Its not just bandwidth though ATi's CEO damit it might have been the CFO don't remember now, also mentioned 32 bit vs 24 bit needed 33% increase in everything from transistors, to number of pixel and vertex shaders to bandwidth.

The fx line didn't have any of those increases compaired to the the 9800."
[the first part is true, he did say that (33% transistor increase to support larger registers and data paths). the second part is not and is not even suggested in the interview (he does not say more PS/VS or bandwidth are necessary). Plus I answered the last part already... of course the NV3x supports FP32 throughout the pipeline. It uses more transistors, just like Nalasco suggests, but it IS in the NV3x/NV40]

 

[cstafs]

The 6800 series gets ~60fps with everything maxed (except AA) at 1600x1200.

I wish they would have done 1280x1024 benchmarks since that's what I have to play at with my LCD.I'm thinking playing at that res. with 2xAA and maybe 4xAF might get me a few more frames then 36,either way I'll enjoy the game when I decide to get it since I'm playing Far Cry now with the AA on high and the AF on medium with everything set to high or very high and it plays nicely even though I don't get much more frames then I would with D3.

 

[Hurin]

Well, I made it to page 22 before giving up.

Some observations:

Claim: Doom3 is just one game!
Yes. But it is one of the most eagerly anticipated games in recent memory. Plus, it will undoubtedly be the engine upon which many other games will be based.

Claim: Doom3 was built with NVIDIA in mind! Cheating!
You've got that backwards, bub. NVIDIA built their newest video cards with Doom3 in mind. And there is nothing wrong with that. When John Carmack spouts off about what he would like to see the next generation of 3D accelerators do, the wise company sits up and takes notice. Because John Carmack's direction can be persuasively argued to be the direction of desktop 3D. Some will claim that ATI chose to ignore Carmack and Doom3 to concentrate on Valve and Half-Life 2. Only to have that blow up in their face. Fine, but you have to concede that it did blow up in their face.

Claim: So what if the x800 is outperformed in Doom3. It's faster everywhere else!
Most reasonable people consider D3D performance and IQ to be essentially tied right now between the flagship products from both companies. When that is taken into account along with these benchmarks from Doom3, it would seem to me that any unbiased person would have to concede that NVIDIA has a superior product with room to grow. While ATI's only hope is wringing more performance out of an already stretched chipset for D3D and a possible OGL-rewrite to close the tremendous gap in Doom3 and other OGL games. Which brings us to. . .

Claim: This doesn't mean crap because ATI is re-writing their OpenGL drivers and that will fix everything!
Fine. They may well be able to make the X800 Pro and XT perform better in OpenGL. But it would be a truly herculean task to bring them up to NVIDIA performance. But, even conceding that a driver rewrite can do that, the ATI person must then concede that such a thing is necessary and that there is therefore some benefit to aligning onself with the proven superior OGL performance and D3D parity of NVIDIA's new cards. One is now a known quantity and is performing brilliantly across the board with future performance in this year's hottest game guaranteed. The other is merely hopeful prayers.

And, for you NVIDIA fans out there:
Claim: This is the death knell for ATI!
This is so ridiculous that it doesn't even warrant a response.

Disclaimer: I had a 9800XT last week. I was waiting for an X800XT-PE to become available. Somehow, while I waited in frustration, I found myself actually looking at the NVIDIA offerings with an open mind. . . and decided to give a 6800GT a try. I think I made the right decision for me even before the Doom3 benches came out. But, honestly, I don't see how anyone can dispute those things said above. In truth, I'd always had disdain for these debates (both sides). But the ATI people really had a chance to be classy here and accept that ATI has taken a major blow. Instead, for the most part, we've just gotten wild conspiracy theories and nonsensical, contradictory arguments ("50fps vs 30fps in D3 doesn't matter. And besides, my X800 is faster in D3D by 3fps!").

 

[Terra]

("50fps vs 30fps in D3 doesn't matter. And besides, my X800 is faster in D3D by 3fps!").

Says it all !

Terra

 

[KakimotoR]

I am either getting 6800Ultra or 6800Ultra Extreme when I am in Taiwan where and when I get my New Year moneys and yes my New Year money normally can afford 2 6800Ultra Extremes + left overs for some Water Cooling Parts

 

[pxc]

I wish they would have done 1280x1024 benchmarks since that's what I have to play at with my LCD.I'm thinking playing at that res. with 2xAA and maybe 4xAF might get me a few more frames then 36,either way I'll enjoy the game when I decide to get it since I'm playing Far Cry now with the AA on high and the AF on medium with everything set to high or very high and it plays nicely even though I don't get much more frames then I would with D3.

You're already going to be running faster than that if your sig is accurate. 560/560 will help get closer to the x800 XT PE framerate than the x800 Pro framerate, even with 4 less pipelines.

http://www2.hardocp.com/article.html?art=NjQyLDM=

at 1024x768 4xaa 8xaf the framerate is 42fps at default clock rate on the x800 Pro. 2xaa, 4xaf will probably be at least in the 50+ fps range for such an overclocked cards at 1280x1024. I wouldn't be surprised if it even did ~55fps in demo1 @ 1280x1024 2xAA 4xAF with a 560/560 clock speed.

 

[oldpablo]

Would have been nice to have seen benchmarks on those systems with the original Doom and Doom2. You know, for comparison.

 

[Dougieha]

Well, I made it to page 22 before giving up.

Some observations:

Claim: Doom3 is just one game!
Yes. But it is one of the most eagerly anticipated games in recent memory. Plus, it will undoubtedly be the engine upon which many other games will be based.

Claim: Doom3 was built with NVIDIA in mind! Cheating!
You've got that backwards, bub. NVIDIA built their newest video cards with Doom3 in mind. And there is nothing wrong with that. When John Carmack spouts off about what he would like to see the next generation of 3D accelerators do, the wise company sits up and takes notice. Because John Carmack's direction can be persuasively argued to be the direction of desktop 3D. Some will claim that ATI chose to ignore Carmack and Doom3 to concentrate on Valve and Half-Life 2. Only to have that blow up in their face. Fine, but you have to concede that it did blow up in their face.

Claim: So what if the x800 is outperformed in Doom3. It's faster everywhere else!
Most reasonable people consider D3D performance and IQ to be essentially tied right now between the flagship products from both companies. When that is taken into account along with these benchmarks from Doom3, it would seem to me that any unbiased person would have to concede that NVIDIA has a superior product with room to grow. While ATI's only hope is wringing more performance out of an already stretched chipset for D3D and a possible OGL-rewrite to close the tremendous gap in Doom3 and other OGL games. Which brings us to. . .

Claim: This doesn't mean crap because ATI is re-writing their OpenGL drivers and that will fix everything!
Fine. They may well be able to make the X800 Pro and XT perform better in OpenGL. But it would be a truly herculean task to bring them up to NVIDIA performance. But, even conceding that a driver rewrite can do that, the ATI person must then concede that such a thing is necessary and that there is therefore some benefit to aligning onself with the proven superior OGL performance and D3D parity of NVIDIA's new cards. One is now a known quantity and is performing brilliantly across the board with future performance in this year's hottest game guaranteed. The other is merely hopeful prayers.

And, for you NVIDIA fans out there:
Claim: This is the death knell for ATI!
This is so ridiculous that it doesn't even warrant a response.

Disclaimer: I had a 9800XT last week. I was waiting for an X800XT-PE to become available. Somehow, while I waited in frustration, I found myself actually looking at the NVIDIA offerings with an open mind. . . and decided to give a 6800GT a try. I think I made the right decision for me even before the Doom3 benches came out. But, honestly, I don't see how anyone can dispute those things said above. In truth, I'd always had disdain for these debates (both sides). But the ATI people really had a chance to be classy here and accept that ATI has taken a major blow. Instead, for the most part, we've just gotten wild conspiracy theories and nonsensical, contradictory arguments ("50fps vs 30fps in D3 doesn't matter. And besides, my X800 is faster in D3D by 3fps!").

Well stated! I agree wholeheartedly.

 

[Ardrid]

Well, I made it to page 22 before giving up.

Some observations:

Claim: Doom3 is just one game!
Yes. But it is one of the most eagerly anticipated games in recent memory. Plus, it will undoubtedly be the engine upon which many other games will be based.

Claim: Doom3 was built with NVIDIA in mind! Cheating!
You've got that backwards, bub. NVIDIA built their newest video cards with Doom3 in mind. And there is nothing wrong with that. When John Carmack spouts off about what he would like to see the next generation of 3D accelerators do, the wise company sits up and takes notice. Because John Carmack's direction can be persuasively argued to be the direction of desktop 3D. Some will claim that ATI chose to ignore Carmack and Doom3 to concentrate on Valve and Half-Life 2. Only to have that blow up in their face. Fine, but you have to concede that it did blow up in their face.

Claim: So what if the x800 is outperformed in Doom3. It's faster everywhere else!
Most reasonable people consider D3D performance and IQ to be essentially tied right now between the flagship products from both companies. When that is taken into account along with these benchmarks from Doom3, it would seem to me that any unbiased person would have to concede that NVIDIA has a superior product with room to grow. While ATI's only hope is wringing more performance out of an already stretched chipset for D3D and a possible OGL-rewrite to close the tremendous gap in Doom3 and other OGL games. Which brings us to. . .

Claim: This doesn't mean crap because ATI is re-writing their OpenGL drivers and that will fix everything!
Fine. They may well be able to make the X800 Pro and XT perform better in OpenGL. But it would be a truly herculean task to bring them up to NVIDIA performance. But, even conceding that a driver rewrite can do that, the ATI person must then concede that such a thing is necessary and that there is therefore some benefit to aligning onself with the proven superior OGL performance and D3D parity of NVIDIA's new cards. One is now a known quantity and is performing brilliantly across the board with future performance in this year's hottest game guaranteed. The other is merely hopeful prayers.

And, for you NVIDIA fans out there:
Claim: This is the death knell for ATI!
This is so ridiculous that it doesn't even warrant a response.

Disclaimer: I had a 9800XT last week. I was waiting for an X800XT-PE to become available. Somehow, while I waited in frustration, I found myself actually looking at the NVIDIA offerings with an open mind. . . and decided to give a 6800GT a try. I think I made the right decision for me even before the Doom3 benches came out. But, honestly, I don't see how anyone can dispute those things said above. In truth, I'd always had disdain for these debates (both sides). But the ATI people really had a chance to be classy here and accept that ATI has taken a major blow. Instead, for the most part, we've just gotten wild conspiracy theories and nonsensical, contradictory arguments ("50fps vs 30fps in D3 doesn't matter. And besides, my X800 is faster in D3D by 3fps!").

Very well said my friend. You read my mind with that post.

 

[pxc]

And, for you NVIDIA fans out there:
Claim: This is the death knell for ATI!
This is so ridiculous that it doesn't even warrant a response.

Yeah, that's just dumb. But of course ATI isn't going anywhere. It still is a serious blow seeing how anticipated D3 is. I read recently that Doom3 is expected to add over $150 million in revenue to Activision. That is huge. And games like this have the effect of enticing people to upgrade their cards.

But as far as video card purchases go, it does seem to have an effect here... undecideds seem to be favoring the 6800GT now. That's not saying much considering how unavailable most of the cards are, unless you want to overpay one of the rip-off vendors (chumbo, Fry's and a few others are the exceptions <3).

 

[cstafs]

You're already going to be running faster than that if your sig is accurate. 560/560 will help get closer to the x800 XT PE framerate than the x800 Pro framerate, even with 4 less pipelines.

http://www2.hardocp.com/article.html?art=NjQyLDM=

at 1024x768 4xaa 8xaf the framerate is 42fps at default clock rate on the x800 Pro. 2xaa, 4xaf will probably be at least in the 50+ fps range for such an overclocked cards at 1280x1024. I wouldn't be surprised if it even did ~55fps in demo1 @ 1280x1024 2xAA 4xAF with a 560/560 clock speed.

Yeah,that's what I was thinking too.I may not be that bad off after all,eh?Thanks for the reply by the way

 

[hamm3rhead]

Anyone paying attention wouldn't be buying a x800pro anyway, they'd buy the x800pro vivo.
Retailing ~$400-$450 16 pipes and 1.6 ns memory. Sure some of them require loading a different firmware, but we are on a hardware site that specializes in tweaking the crap out of things and voiding warranties.

 

[burningrave101]

But as far as video card purchases go, it does seem to have an effect here... undecideds seem to be favoring the 6800GT now. That's not saying much considering how unavailable most of the cards are, unless you want to overpay one of the rip-off vendors (chumbo, Fry's and a few others are the exceptions <3).

The 6800GT and 6800nu are readily available at quite a few places and more are showing up every day. The 6800u's have been shipping lately to. PNY just shipped out a bunch more and so did XFX. Gainward is shipping theirs.

The only real vaporware right now is the X800XT PE. Have you seen the Gateway threads? Gateway says they aren't going to ship any more X800XT's and their offering them the X800Pro instead lol. And CDW isn't suppost to get in their shipment till 2010.

 

[Dijonase]

Well, like I said earlier, the X800s may be good cards, but seeing as how the 6800GT is vastly superior in Doom3 (or any OpenGL title for that matter) and pretty much dead even (or slightly ahead or behind) in D3D games, then I have no problems picking up a 6800GT from eVGA and selling my X800 Pro on ebay. The lowest I see them going for on ebay is $375 or so anyway, so I should get most - if not all - of my money back.

 

[GoCanes]

Maybe they used 2GB and 4GB RAM systems to remove possible bottlenecks to allow the cards to go head to head instead of throttling performance...

 

[Wrench00]

The 6800GT and 6800nu are readily available at quite a few places and more are showing up every day. The 6800u's have been shipping lately to. PNY just shipped out a bunch more and so did XFX. Gainward is shipping theirs.

Not in Canada