GF100 spotted?

[Elledan]

i suggest giving this a read, a great bit of info in there on why DX11 will become a standard.

http://www.bit-tech.net/bits/interviews/2010/01/06/interview-amd-on-game-development-and-dx11/1

specifically



this has to be great news for developers and should make DX11 really help its uptake. You don't need to build a DX11 exclusive game, but if you do build it in DX11 it automatically works on DX10 and DX9 hardware.

So technically speaking no game will ever be DX11-exclusive then as DX11 is the 9 through 11 runtimes rolled together into one package

 

[Unknown-One]

So technically speaking no game will ever be DX11-exclusive then as DX11 is the 9 through 11 runtimes rolled together into one package

They will be, but thanks to properties inherent to DX11, they will also run on older hardware with reduced functionality.


I'm really interested in Nvidia's driver update to allow triple-monitor on older cards. Even if it requires SLI, grabbing a second GTX260 doesn't sound like such a bad idea, especially if you look at the cost of current ATi cards.

An HD5850 is $300 (+$100 DisplayPort to DVI adapter). Now, if you already have one GTX260, then it only costs $185 to get a second one. Not only that, but if the benchmarks are to be believed, then two GTX260's in SLI appear to be faster than a single HD5850 in all cases, and faster than an HD5870 in a few select scenarios.

I'm first in line for a second card if Nvidia gets these drivers out. I'm already running triple-monitor with SoftTH, but this should perform significantly better

 

[Xzyrus]

They will be, but thanks to properties inherent to DX11, they will also run on older hardware with reduced functionality.

Ahh I was thinking more clear cut advancement where old DX9-10.1 wouldn't be able to render, but I guess they thought DX11 out well. I was thinking it'd require active parallel coding in the different DX versions. Technically speaking then DX9 and 10.1 card owners have nothing to worry about then. Makes me think I should have picked up a 4850 when it was at EOL.

 

[Angus66]

Wow I've never seen such a detailed wood carving of a video card before. lol I kid, but seriously, pictures of this card keep getting "leaked" and months from release. I for one am sick of looking at it, nvidia needs to put up or shut up.

+10000

 

[heatlesssun]

They will be, but thanks to properties inherent to DX11, they will also run on older hardware with reduced functionality.


I'm really interested in Nvidia's driver update to allow triple-monitor on older cards. Even if it requires SLI, grabbing a second GTX260 doesn't sound like such a bad idea, especially if you look at the cost of current ATi cards.

An HD5850 is $300 (+$100 DisplayPort to DVI adapter). Now, if you already have one GTX260, then it only costs $185 to get a second one. Not only that, but if the benchmarks are to be believed, then two GTX260's in SLI appear to be faster than a single HD5850 in all cases, and faster than an HD5870 in a few select scenarios.

I'm first in line for a second card if Nvidia gets these drivers out. I'm already running triple-monitor with SoftTH, but this should perform significantly better

I totally agree. I already have three DVI/HDMI 24" 1920x1200 monitors and 3 280s. If "nFinity" works as well or better as Eyefinity on GTX 2XXs, then nVidia will gain a lot of rep with me. A single 280 can still crush most games at max at 4AA+ at 1920x1200, 3 280's should be able to do the same at 5760x1200. The only thing that would be missing is DX 11 which so far hasn't shown me much. I played Dirt 2 on a 5770 system about a month ago and DX 11 just didn't show me anything over DX 9 as I have the post-processing effects turned on for non DX11 hardware.

 

[sweetreddyd]

My GTX 260s were very, very sad at 2560x1600, and I had tri-sli. I gave two to my brother and sold one for 80$.

Thats about how comprable 260s are to 5870s at high resolutions.

 

[heatlesssun]

My GTX 260s were very, very sad at 2560x1600, and I had tri-sli. I gave two to my brother and sold one for 80$.

Thats about how comprable 260s are to 5870s at high resolutions.

3 260s couldn't drive what at 2560x1600. For games that support SLI, 3 260s should have now problem at 2560x1600. I've never had my 280s at that resolution but everything I've read indicates 2560x1600 would be no sweat most of the time.

 

[sweetreddyd]

Name the modern game, and they struggled. Batman, frequently under 60, impossible to use AA in any reasonable sense.

Shift, impossible to use AA without serious alpha blending slowdowns.

Burnout paradise - SSAO completely destroyed framerate, cutting it in half.

Clear Sky - Impossible to max settings without seeing 30fps average.

Borderlands - Impossible to run at 60fps without disabling shadows and SSAO, terrible slowdowns when using brick, rendering melee useless any time when particle effects were covering more than 30 - 40% of screen space.

Mechwarrior living legends mod - Basically a nightmare to play anytime fire or overlapping transparency occurred.

Ghostbusters - Terrible framerates when fill rate was taxed with particle streams overlapping, when the framerate in the game dips below 40, very hard to control, very bad for playability.

Fallout 3 - Freezes of up to 10 seconds after activating VATS with ANY anti-aliasing enabled, awful, awful performance when actually in VATS.

GRID - Awful framerates in japanese tracks, worse with reflection heavy maps.

RE5 - Impossible to run with any antialiasing and maintain 60fps, driver bug after 190.62 negatively impacting SLI scaling on water level.

Wolverine Game - Impossible to maintain 60fps at 2560 with any AA level, including 2x.

SF IV - Great framerates in benchmarks, terrible framerates on cruise ship level with antialiasing enabled.

Thats the fallacy in reading instead of owning I suppose. Mileage may vary, but I was displeased enough with the high res performance to dump them all for 5870s, even after one of my new 5870s was bad out of the box.

The 260s did NOT scale well when I moved up from 1920x1200 to 2560x1600. Poor minimum framerates, and slowdowns at all the inconvenient times that inspire people to upgrade video cards for. nfinity I believe will likely exhibit the exact same scaling problems on gtx 260s. Simply not enough framebuffer for an acceptable experience at the highest resolutions and highest settings.

280s and 285s are a different story.

~1080p and 1600p+(and thus eye/nfinity resolutions) are wildly different situations, while it's very easy to assume/believe performance will scale well, I found it not to be the case in reality, and ended up spending a lot of money to improve performance to the level I find acceptable.

A second, or third gtx 260 didn't really cut it for me at 2560 with antialiasing, and the recent eyefinity performance review paints a pretty clear picture that at some of these multiscreen resolutions, even 5870s, which are twice the speed plus some per card, are struggling quite a bit to maintain playability, let alone a consistent 60fps goal.

In short - triple GTX 260s couldn't hack high res and antialiasing with max settings in most games I play, and maintain 60fps, with even higher resolutions, it's going to get even worse, higher resolutions play to the 260s weaknesses, not it's strengths.

And just in case anyone wants to 'lol u didn't have triple gtx 260s' - http://service.futuremark.com/compare?3dmv=1027967

My new rig gets the same score, only it handles actual video games, it's not a benchmark queen card that can't handle rendering smoke and fire in dragon age without tanking to 20 fps.

 

[Unknown-One]

My GTX 260s were very, very sad at 2560x1600, and I had tri-sli. I gave two to my brother and sold one for 80$.

Thats about how comprable 260s are to 5870s at high resolutions.

Name the modern game, and they struggled. Batman, frequently under 60, impossible to use AA in any reasonable sense.

Shift, impossible to use AA without serious alpha blending slowdowns.

Mechwarrior living legends mod - Basically a nightmare to play anytime fire or overlapping transparency occurred.

Ghostbusters - Terrible framerates when fill rate was taxed with particle streams overlapping, when the framerate in the game dips below 40, very hard to control, very bad for playability.

Fallout 3 - Freezes of up to 10 seconds after activating VATS with ANY anti-aliasing enabled, awful, awful performance when actually in VATS.

GRID - Awful framerates in japanese tracks, worse with reflection heavy maps.

(Hope you don't mind, I chopped out the games I don't have from the list to shorten the quote.)

Your numbers aren't making any sense to me. I've got a single GTX260, and a 9800GT dedicated to PhysX, and I'm getting better performance than your tri-SLI setup? I don't believe it. The figures you're posting make it sound like you didn't even have SLI enabled. Take a look at what I'm getting with a single card:

Batman Arkham Asylum:
5040x1050, max settings, no AA, average framrate 42 (That was with PhysX on full)

NFS Shift:
5040x1050, max settings, no AA, average framrate 46

Mechwarrior Living Legends (Crysis Mod):
5040x1050, medium settings, EdgeAA, average framrate 58

Ghostbusters:
5040x1050, max settings, no AA, average framrate 38 (too slow, usually run at 1680x1050)

Fallout 3:
5040x1050, max settings, no AA, average framrate 46

GRID:
5040x1050, max settings, 4xAA, average framrate 49

Thats the fallacy in reading instead of owning I suppose. Mileage may vary, but I was displeased enough with the high res performance to dump them all for 5870s, even after one of my new 5870s was bad out of the box.

The 260s did NOT scale well when I moved up from 1920x1200 to 2560x1600. Poor minimum framerates, and slowdowns at all the inconvenient times that inspire people to upgrade video cards for. nfinity I believe will likely exhibit the exact same scaling problems on gtx 260s. Simply not enough framebuffer for an acceptable experience at the highest resolutions and highest settings.

280s and 285s are a different story.

Now I know you're full of it, the GTX280 isn't that much faster than the 260, much less three of them in SLI. A quick glance at the benchmarks confirms it. GTX260 tri-SLI keeps up with two GTX280's in SLI just fine.

It's about a 50/50 split. Sometimes the three GTX260's come out way ahead, other times they're a little bit behind the two GTX280's. That's to be expected, each setup has different strengths; the three GTX260's have more total stream processors, while the two GTX280's have a bit more video RAM and bandwidth, and don't have as much CPU overhead (no 3rd card). That said, both solutions are more than fast enough to run games at 2560x1600 (with the possible exception of Crysis.)

~1080p and 1600p+(and thus eye/nfinity resolutions) are wildly different situations, while it's very easy to assume/believe performance will scale well, I found it not to be the case in reality, and ended up spending a lot of money to improve performance to the level I find acceptable.

A second, or third gtx 260 didn't really cut it for me at 2560 with antialiasing, and the recent eyefinity performance review paints a pretty clear picture that at some of these multiscreen resolutions, even 5870s, which are twice the speed plus some per card, are struggling quite a bit to maintain playability, let alone a consistent 60fps goal.

In short - triple GTX 260s couldn't hack high res and antialiasing with max settings in most games I play, and maintain 60fps, with even higher resolutions, it's going to get even worse, higher resolutions play to the 260s weaknesses, not it's strengths.

And just in case anyone wants to 'lol u didn't have triple gtx 260s' - http://service.futuremark.com/compare?3dmv=1027967

My new rig gets the same score, only it handles actual video games, it's not a benchmark queen card that can't handle rendering smoke and fire in dragon age without tanking to 20 fps.

Something is SERIOUSLY wrong with your setup. I'm currently gaming at 5040x1050 with a single GTX260 without too much trouble (5040x1050 is larger than 2560x1600, for reference). I'm also dealing with the overhead added by SoftTH copying a 1680x1050 frame over the PCIe bus 60 times every second.

I'm getting decent performance ONE card, I don't see how you could possibly be experiencing such horrible performance with THREE in SLI.

 

[realitycheck]

Name the modern game, and they struggled. Batman, frequently under 60, impossible to use AA in any reasonable sense.

Shift, impossible to use AA without serious alpha blending slowdowns.

Burnout paradise - SSAO completely destroyed framerate, cutting it in half.

Clear Sky - Impossible to max settings without seeing 30fps average.

Borderlands - Impossible to run at 60fps without disabling shadows and SSAO, terrible slowdowns when using brick, rendering melee useless any time when particle effects were covering more than 30 - 40% of screen space.

Mechwarrior living legends mod - Basically a nightmare to play anytime fire or overlapping transparency occurred.

Ghostbusters - Terrible framerates when fill rate was taxed with particle streams overlapping, when the framerate in the game dips below 40, very hard to control, very bad for playability.

Fallout 3 - Freezes of up to 10 seconds after activating VATS with ANY anti-aliasing enabled, awful, awful performance when actually in VATS.

GRID - Awful framerates in japanese tracks, worse with reflection heavy maps.

RE5 - Impossible to run with any antialiasing and maintain 60fps, driver bug after 190.62 negatively impacting SLI scaling on water level.

Wolverine Game - Impossible to maintain 60fps at 2560 with any AA level, including 2x.

SF IV - Great framerates in benchmarks, terrible framerates on cruise ship level with antialiasing enabled.

Thats the fallacy in reading instead of owning I suppose. Mileage may vary, but I was displeased enough with the high res performance to dump them all for 5870s, even after one of my new 5870s was bad out of the box.

The 260s did NOT scale well when I moved up from 1920x1200 to 2560x1600. Poor minimum framerates, and slowdowns at all the inconvenient times that inspire people to upgrade video cards for. nfinity I believe will likely exhibit the exact same scaling problems on gtx 260s. Simply not enough framebuffer for an acceptable experience at the highest resolutions and highest settings.

280s and 285s are a different story.

~1080p and 1600p+(and thus eye/nfinity resolutions) are wildly different situations, while it's very easy to assume/believe performance will scale well, I found it not to be the case in reality, and ended up spending a lot of money to improve performance to the level I find acceptable.

A second, or third gtx 260 didn't really cut it for me at 2560 with antialiasing, and the recent eyefinity performance review paints a pretty clear picture that at some of these multiscreen resolutions, even 5870s, which are twice the speed plus some per card, are struggling quite a bit to maintain playability, let alone a consistent 60fps goal.

In short - triple GTX 260s couldn't hack high res and antialiasing with max settings in most games I play, and maintain 60fps, with even higher resolutions, it's going to get even worse, higher resolutions play to the 260s weaknesses, not it's strengths.

And just in case anyone wants to 'lol u didn't have triple gtx 260s' - http://service.futuremark.com/compare?3dmv=1027967

My new rig gets the same score, only it handles actual video games, it's not a benchmark queen card that can't handle rendering smoke and fire in dragon age without tanking to 20 fps.

I'm going to have to agree with the other guy who quoted you. Either your system wasn't set up correctly or you're fibbing. My sli 260's owned the piss out of everything, at every res, except crysis.

I might actually sell my 4850 and pick up two 260's if Nvidia release nfinity drivers.

 

[sweetreddyd]

Yes, hmm, because my 30014 3dmark score isn't consistent with tri-sli setups. Maybe check the evga 260 tri-sli benchmark thread, 30k is smack in the middle.

45 fps averages might be acceptable playability to you, but they aren't to me, because they mean frequent dips into minimums under 30. I have no issues forcing 3-frame AFR, either, but thanks for being condescending without actually having the hardware yourself(which means a lot of internets learning, gg).

Go ahead and believe your SLI gtx 260s will be fine in 'nfinity'. The fact remains, 5870s destroy a 260 in every conceivable benchmark, and in the eyefinity performance eval on this very site, you can see excessive hard dips below 60 fps, which to me, is unplayable.

I'll wait for your 'omg hep nfinity performance bad' thread in 3 months.

So, basically, what you did, was post similar framerates for your rig to what I was stating(well under 60 fps average), the same kind of framerates that motivated me to upgrade, and said they were good.

Notice I never mention 'average framerate of 45' I say 60fps. 60 frames per second is my playability goal, not the ugly inconsistent framerate that an average of 45 implies.

Fermi I have no doubt will be able to muster decent framerates at nfinity resolutions. however using your 260s for it, will likely motivate you to upgrade away from them, and instill a great amount of buyers remorse if you rely on them to provide a good gaming experience without turning the resolution down.

And if you honestly think 285s and 280s 'aren't that much faster' at the higher resolutions, you are fooling yourself. 260s scale fine on a 1050 screen, they do NOT at higher resolutions, and you are posting secondhand info, wheras I owned them and found them totally unacceptable because of just the type of framerate variability you've just posted.

This is the problem that occurs when you fish the internet for benchmarks and don't own the hardware yourself and acquire firsthand knowledge of it's limitations. You think you are an expert when you have one of the cards in question. I had 3, and a 2560 screen. You have one, and a 1680x1050 monitor. And you're telling me I'm full of it because of benchmarks you saw on the internet. Not at all comprable.

What internet benchmarks (not including the ones on [H] don't tell you, is how often or in what places the minimum framerate dips occurred. In this situation benchmarks someone else did are no substitute for actual working experience with the hardware.

But you're going to go on thinking they are, and thats fine.

 

[Unknown-One]

45 fps averages might be acceptable playability to you, but they aren't to me, because they mean frequent dips into minimums under 30.
[...]
So, basically, what you did, was post similar framerates for your rig to what I was stating(well under 60 fps average), the same kind of framerates that motivated me to upgrade, and said they were good.

I'm going to stop you right there.

My numbers were form a SINGLE CARD. You're supposedly running THREE OF THEM. You should be seeing performance far in excess of my numbers. Benchmarks from other website agree that three GTX260's should perform a hell of a lot faster than one.

2560x1600 is actually lower resolution that 5040x1050, so that should be helping you out as well...

This is the problem that occurs when you fish the internet for benchmarks and don't own the hardware yourself and acquire firsthand knowledge of it's limitations. You think you are an expert when you have ONE of the cards in question.

Yes, because one card is fairly decent for 5040x1050, so a second one should improve things. Makes sense to me.

I had 3, and a 2560 screen. You have one, and a 1680x1050 monitor

No, I have one GTX260 and THREE 1680x1050 monitors, for a combined horizontal resolution of 5040x1050.

You say you were running three GTX260's on a smaller 2560x1600 display and getting worse performance. Something doesn't add up on your end. 3x the graphics cards with a smaller monitor should be faster, UNLESS there's something else wrong with your system (maybe you're CPU limited?).

 

[sweetreddyd]

And you're assuming you know how 3 cards scale based on your experience with them, which is reading on the internet, VS testing yourself, thus, the fallacy.

 

[Unknown-One]

And you're assuming you know how 3 cards scale based on your experience with them, which is reading on the internet, VS testing yourself, thus, the fallacy.

I'm assuming that three GTX260's will be faster than one, that is all. According to your numbers, three GTX260's are slower than one. Doesn't make any sense at all, so I say you're full of it.

 

[sweetreddyd]

Not slower, but not appreciably faster when dealing with transperancy or fillrate heavy scenes. And I already said, my goal is 60fps, wheras yours seems to be an EXTREMELY CHOPPY average of 45.

You are assuming you know how additional cards scale based on benchmarks which show you average framerates and make no mention of the dips below minimum playable.

 

[Unknown-One]

Not slower, but not appreciably faster when dealing with transperancy or fillrate heavy scenes. And I already said, my goal is 60fps, wheras yours seems to be an EXTREMELY CHOPPY average of 45.

First off, 45 isn't all that choppy. Second, who said that was my goal? Of course my goal is 60FPS, why else would I be looking at adding another card?

You are assuming you know how additional cards scale based on benchmarks which show you average framerates and make no mention of the dips below minimum playable.

I assume three will be faster than one. If you're honestly saying that three are the same or slower than one, then i can't believe you.

 

[sweetreddyd]

I'm saying the poor minimum framerate, which is the annoying reason I upgrade video cards for, is not appreciably better between 3 260s in AFR 3. There is a nuance here that you don't understand, and I'm done trying to educate you about it.

You can quote benchmarks (you got from other people or googled) as much as you want, never having posessed the hardware, and been annoyed by frequent dips below 30 frames at high resoltions, I wouldn't expect you to understand. But the fact remains I gave dual 260s away for free, and sold one for 80$ I was so displeased with high resolution performance with AA enabled. And my 3dmark score called, and it says there was nothing wrong with my setup at all.

 

[Unknown-One]

Come on man, the minimum framrates can't be that bad. [H]ard|OCP themselves show GTX260 SLI (not tri-SLI) playing Fallout 3 at 2560x1600 with a 35FPS minimum with 16x csAA and 16x AF.

They also show GTX260 SLI running COD5 at 2560x1600 with 2x SSAA, which means it was internally rendering at 51200x3200 and scaling it down to 2560x1600. It still managed to maintain 35FPS minimum while rendering 4x more pixels than required.

I'm betting you could easily get the minimum in both of those games over 60 if you used something more sane like 4x AA.

 

[sweetreddyd]

But what benchmarks, even [H] benchmarks don't show you, is the odd time you hit 'V' in fallout 3 while looking at a raider in front of a window, with water in the background, where it takes nearly 10 seconds to even start the VATS animation, with anything over 2x AA forced on, and all of a sudden you are getting 15fps, and then go shopping for a new video card.

Thats why benchmarks are all well and good, but not a concrete substitute for actual real world gaming. It's just like when quake 3 was the defacto benchmark, and you could get 100 fps at 1600x1200, but the second you are zoomed into a rocket trail, your framerate is 25. Anything under 30 and I'm out buying hardware.

But I'm tired of derailing this thread over it, suffice to say, if you don't mind miniumum framerates touching 20ish, you probobly will be just fine with 260s driving multimonitor, I personaly can't stand anything under 60.

 

[realitycheck]

Yes, hmm, because my 30014 3dmark score isn't consistent with tri-sli setups. Maybe check the evga 260 tri-sli benchmark thread, 30k is smack in the middle.

45 fps averages might be acceptable playability to you, but they aren't to me, because they mean frequent dips into minimums under 30. I have no issues forcing 3-frame AFR, either, but thanks for being condescending without actually having the hardware yourself(which means a lot of internets learning, gg).

Go ahead and believe your SLI gtx 260s will be fine in 'nfinity'. The fact remains, 5870s destroy a 260 in every conceivable benchmark, and in the eyefinity performance eval on this very site, you can see excessive hard dips below 60 fps, which to me, is unplayable.

I'll wait for your 'omg hep nfinity performance bad' thread in 3 months.

So, basically, what you did, was post similar framerates for your rig to what I was stating(well under 60 fps average), the same kind of framerates that motivated me to upgrade, and said they were good.

Notice I never mention 'average framerate of 45' I say 60fps. 60 frames per second is my playability goal, not the ugly inconsistent framerate that an average of 45 implies.

Fermi I have no doubt will be able to muster decent framerates at nfinity resolutions. however using your 260s for it, will likely motivate you to upgrade away from them, and instill a great amount of buyers remorse if you rely on them to provide a good gaming experience without turning the resolution down.

And if you honestly think 285s and 280s 'aren't that much faster' at the higher resolutions, you are fooling yourself. 260s scale fine on a 1050 screen, they do NOT at higher resolutions, and you are posting secondhand info, wheras I owned them and found them totally unacceptable because of just the type of framerate variability you've just posted.

This is the problem that occurs when you fish the internet for benchmarks and don't own the hardware yourself and acquire firsthand knowledge of it's limitations. You think you are an expert when you have one of the cards in question. I had 3, and a 2560 screen. You have one, and a 1680x1050 monitor. And you're telling me I'm full of it because of benchmarks you saw on the internet. Not at all comprable.

What internet benchmarks (not including the ones on [H] don't tell you, is how often or in what places the minimum framerate dips occurred. In this situation benchmarks someone else did are no substitute for actual working experience with the hardware.

But you're going to go on thinking they are, and thats fine.

Dude, calm down. I'm just saying my sli setup got way over 60 fps in every game except crysis, with some small exceptions here and there.

It might work, it might not.

And of course a 5870 destroys a single gtx260. But most benchmarks show sli gtx 260's beating a single 5870, so how do you think tri-sli, let alone regular 260's won't handle nfinity.

Just saying.. I can get your some horse tranquilizers too if you want.

 

[Unknown-One]

But what benchmarks, even [H] benchmarks don't show you, is the odd time you hit 'V' in fallout 3 while looking at a raider in front of a window, with water in the background, where it takes nearly 10 seconds to even start the VATS animation, with anything over 2x AA forced on, and all of a sudden you are getting 15fps, and then go shopping for a new video card.

You might want to patch your game, I haven't seen the VATS bug since release day. If you're still experiencing it, then there's something else wrong with your system.

But I'm tired of derailing this thread over it, suffice to say, if you don't mind miniumum framerates touching 20ish, you probobly will be just fine with 260s driving multimonitor, I personaly can't stand anything under 60.

I don't think I'll have any trouble keeping things over 60FPS, thank you very much. Benchmarks agree with me, including the ones that show minimum framrate and full framrate graphs like [H]'s. I see realitycheck, who HAS had experience with the hardware, is also getting better framrates than you (and agrees with two GTX260's being faster than an HD5850).

I don't care what your 3DMark score is, it sounds like your hardware wasn't configured properly.

 

[heatlesssun]

Well I expect that my 3 280s will be able to easily beat any single AMD card solution solution even now on the market, including the 5970 with multi-monitor. As Unknown-One and realitycheck have stated, I think that performance would be consistent with single monitor benchmarks assuming that multi-monitor monitor performance scales as one would logically expect but obviously without the drivers this is all speculation.

 

[sweetreddyd]

Right, because I'm not using the exact same hardware with triple 5870s, and you need to manually patch the steam version of fallout 3.

You are looking at the world through fanboy glasses, and I really don't buy those numbers with a crapstick low OCed phenom. I wouldn't be getting 30k in vantage if my system wasn't configured properly.

And again - we aren't discussing your 3 280s, we are discussing two GTX 260s in SLI, a setup which I have direct experience with, and found lacking.

But hey, keep thinking what you think, I have 3 5870s and a 30 inch monitor, I haven't seen sub 60 fps in weeks.

 

[Unknown-One]

You are looking at the world through fanboy glasses

No, we're looking at your numbers, looking at our numbers, comparing with benchmarks, cross-referencing, and coming to the conclusion that your numbers don't make much sense.

I really don't buy those numbers with a crapstick low OCed phenom.

You call us fanboys, then break out with that? I think that makes you sound like a massive intel fanboy, the hypocrisy is so thick in the air it's almost palpable

My Phenom II X4 is plenty fast. At least, fast enough not to be CPU limited. My graphics card, on the other hand, is quite taxed at 5040x1050. An SLI solution should offer a nice little boost.

I wouldn't be getting 30k in vantage if my system wasn't configured properly.

I've seen it done. Bad driver that worked fine with 3DMark but tanked performance in a number of other things.

 

[sweetreddyd]

You could call me a fanboy. A fanboy of 2560x1600 locked at 60fps. If a phenom were somehow capable of maintaining a constant framerate at that speed, I'd probobly have one.

Instead I have a decently overclocked I7. Your numbers and my numbers don't match up, likely because yours are made up, and to this point, I'm the only one who has posted a link to an actual benchmark, but you don't like that one.

And so far, you haven't actually refuted anything I've posted in any acceptable way, in fact, your garbage estimated numbers at your triplehead 1050 res pretty much support what I'm saying. If you're getting an average of 45, you're seeing the kind of minimums that drove me to buy new hardware, you just don't mind crappy framerates. And as I said, that is perfectly ok. I'm not content with anything under 60. My system is fine, 260s are just not as fast as you seemingly wish they were.

 

[Unknown-One]

Your numbers and my numbers don't match up, likely because yours are made up, and to this point, I'm the only one who has posted a link to an actual benchmark, but you don't like that one.

I thought you said benchmarks were worthless and didn't show the true nature of real-world gameplay? Now you're falling back on one?

I assure you, my numbers are real. I'm not the only one here telling you that your numbers don't add up. There are people here posting with GTX260 SLI setups that have attested to their performance.

And so far, you haven't actually refuted anything I've posted in any acceptable way, in fact, your garbage estimated numbers at your triplehead 1050 res pretty much support what I'm saying.

Wait, you just said you thought my numbers were made up. Now you're using them to support your own claims? That doesn't make any sense... they either are real or they aren't, make up your damn mind.

If you're getting an average of 45, you're seeing the kind of minimums that drove me to buy new hardware

Yes, i'm seeing them on ONE CARD. You had THREE of them. You should have been seeing much faster framrates than you say you were.

you just don't mind crappy framerates. And as I said, that is perfectly ok. I'm not content with anything under 60. My system is fine, 260s are just not as fast as you seemingly wish they were.

I do mind crappy framrates, otherwise I wouldn't be looking at a second GTX260 for SLI, now would I?

And your system is most certainly NOT fine if your performance was nerfed so heavily that my single GTX260 was performing the same as your tri-SLI setup with THREE of the darn things.

 

[sweetreddyd]

I never said it performed the same, I said minimum framerates in bad situations at high resolutions were poor enough to motivate me to upgrade. If there were a problem with my system the exact same system with 3 ati cards would not be running flawlessly.

The problem is with GTX 260s at high resolutions. Go ahead, buy another one. Enjoy your bad framerates. You're already getting what I considered to be unplayable framerates. I'd suggest you go back and read what I actually posted to get a better grasp on what my message to you is.

No actually, no I wouldn't. No. On second thought, I suggest you go out and buy a second 260 and wow people with your average framerates which still hover in the 50ish at your resolution. The thing is, you basically made up a bunch of gaming numbers, which support exactly what I'm saying anyway (and you would realize this I think, if you rationally examined my posts, and used cognition in any way) and then railed on about how wrong I was. I don't accept framerates under 60, 260s could not push 2560 and maintain 60fps, not even 3 of them, in intense situations. This is the point you don't seem to understand, and won't most likely, until you make your misguided purchase thinking you are buying performance you simply are not.

Enjoy that. I'm starting to realize I should have sold YOU my 260s.

 

[kadozer]

sweetreddyd 10, google guy 0.

 

[Nvidia101]

they look to be the same size as the gtx 295 card.

 

[shansoft]

Dude, calm down. I'm just saying my sli setup got way over 60 fps in every game except crysis, with some small exceptions here and there.

It might work, it might not.

And of course a 5870 destroys a single gtx260. But most benchmarks show sli gtx 260's beating a single 5870, so how do you think tri-sli, let alone regular 260's won't handle nfinity.

Just saying.. I can get your some horse tranquilizers too if you want.

I can tell someone is smoking weed here..

I had a GTX 295, and the fps number were high on max fps, but min fps were terrible in most game I have play, especially Crysis and Borderlands.

After switching to 5870, the smoothness of the game are generally more smoother. since GTX 295 and GTX 260 SLI are similar, there is no way GTX 260 SLI can outperform 5870.

also, I have a 5970 right now and I am totally regret having it when it have worse min fps than my 5870 in some game.

 

[Decko]

Okay I didn't read the whole post, but can I ask an honest question? What about GTX 260 tri-sli causes such terrible minimum frame rate? And why would a 5970 cause lower frame rates than a 5970?

 

[shansoft]

Okay I didn't read the whole post, but can I ask an honest question? What about GTX 260 tri-sli causes such terrible minimum frame rate? And why would a 5970 cause lower frame rates than a 5970?

driver issue? game engine issue?

don't know..... but it happens....

 

[realitycheck]

I never said it performed the same, I said minimum framerates in bad situations at high resolutions were poor enough to motivate me to upgrade. If there were a problem with my system the exact same system with 3 ati cards would not be running flawlessly.

The problem is with GTX 260s at high resolutions. Go ahead, buy another one. Enjoy your bad framerates. You're already getting what I considered to be unplayable framerates. I'd suggest you go back and read what I actually posted to get a better grasp on what my message to you is.

No actually, no I wouldn't. No. On second thought, I suggest you go out and buy a second 260 and wow people with your average framerates which still hover in the 50ish at your resolution. The thing is, you basically made up a bunch of gaming numbers, which support exactly what I'm saying anyway (and you would realize this I think, if you rationally examined my posts, and used cognition in any way) and then railed on about how wrong I was. I don't accept framerates under 60, 260s could not push 2560 and maintain 60fps, not even 3 of them, in intense situations. This is the point you don't seem to understand, and won't most likely, until you make your misguided purchase thinking you are buying performance you simply are not.

Enjoy that. I'm starting to realize I should have sold YOU my 260s.

Dude, you are getting owned, adn it's because of your attitude. Yeah, you have had experience with your 260's, which, if you aren't fibbing, had something wrong with them.

Other benchmarks, my personal experience, and others' personal experience don't match up with yours.

But it doesn't really matter. IF Nvidia makes their multiple monitor solution available to the 260's, then we will see. Until then, the argument is all speculation. In fact, I think the only reason it's an argument is because of you. Instead of sayinig, "hey, these are the numbers I got, and I wasn't happy with them, maybe something was amiss", instead of, "fuck all of you and your shitty 260's you plebes, I'm a king and I wipe my ass with 260's because they only get 3 fps in Doom while my 5870 gets 267 fps in Crysis 2!!!!1!!1!!", maybe you wouldn't be getting spanked right now.

 

[maximus87]

I see sweet's point, but he sure is making it poorly. You will see higher highs with 260 SLI/Tri-SLI setups, but the lows are still the same in certain circumstances. For some it's a problem (obviously it was for him) others likely won't notice it.

 

[Kaldskryke]

"fuck all of you and your shitty 260's you plebes, I'm a king and I wipe my ass with 260's because they only get 3 fps in Doom while my 5870 gets 267 fps in Crysis 2!!!!1!!1!!"

I lol'd

But this thread is seriously derailed - is there any way we can get OT? I came here looking for news about GF100 and found this argument about Tri-Sli 260 instead... If people want to explicitly discuss how well 3D-Surround will work on their current hardware, maybe make a thread for it.

 

[sweetreddyd]

Dude, you are getting owned, adn it's because of your attitude. Yeah, you have had experience with your 260's, which, if you aren't fibbing, had something wrong with them.

Other benchmarks, my personal experience, and others' personal experience don't match up with yours.

But it doesn't really matter. IF Nvidia makes their multiple monitor solution available to the 260's, then we will see. Until then, the argument is all speculation. In fact, I think the only reason it's an argument is because of you. Instead of sayinig, "hey, these are the numbers I got, and I wasn't happy with them, maybe something was amiss", instead of, "fuck all of you and your shitty 260's you plebes, I'm a king and I wipe my ass with 260's because they only get 3 fps in Doom while my 5870 gets 267 fps in Crysis 2!!!!1!!1!!", maybe you wouldn't be getting spanked right now.

You are making the mistake of thinking there is an internet army behind you, when you actually have no clue what you are talking about.

I think the score was called at 10 me actually, which would be you getting owned, because you don't know what you are talking about. You have 280s, which at high resolutions are a completely different story than 260s, most especially at the resolutions we are discussing with an nfinity driver update, or 2560x1600. Maybe if you understood the concept of a minimum framerate, you could have accepted what I said as what it is, REALITY.

If a single card cannot render a frame fast enough to display 60fps, why would three cards render the same frame fast enough to display 60fps? When each card is identical and they are not working together AT ALL to render each individual frame. Thats right, you heard it here first, AFR is called ALTERNATE FRAME RENDERING for a reason.

Again, I'm not saying 260s are a terrible card, I'm saying they give terrible minimum framerates at resolutions over 1080p with any kind of antialiasing enabled. But the cognitive dissonance here is quite extreme.

Fermi will be able to handle maintaining a 60hz refresh at some of the higher resolutions, of this I have no doubt, but a card that is half the speed of a 5870, with fermi being reportedly 20 - 30% over that? Good luck with that.

When I had 260s mated to a 1200p screen, I had no problem with them, however, I bought a much bigger screen, and noticed that in AFR modes, scenes that would struggle greatly, and guess what, per frame, alternate frame rendering makes no difference at all.I So I suggest you check out some of the benchmarks online you seem so fond of, that show dual sli 280s VS tri-sli 260s. It might be an unpleasant surprise for you,too bad you never bothered to research what you are talking about, like I did in a hands on environment, and you've done only in imaginationland.

You carry the true hallmark of a fanboy, getting upset when someone else says the setup they like didn't perform to expectation, and defaulting to insulting that person and their rig, rather than rationally examine and research the subject.

The reality of this portion of the thread, is you have someone with a single 260, assuming he knows what kind of minimum framerate he'll get based on what the internet says about average frames, and someone with tri-sli 280s, who thinks he knows as well, and if you'd bother to look at some of the comparisons between tri-sli 260s, and even normal sli 280s, you'd know they scale quite a bit differently at high resolutions.

I'm thinking they had SLI fermis powering the 3d multimonitor setups for a very good reason. Even the 5970 struggles to put up 60 frames on those resolutions, but yeah, I'm sure the SLI 260s will deliver a great experience at that res, and have no problem getting 60 frames. derp.

 

[heatlesssun]

All I know is that a single 280 is still a hell of a card. SLI isn't supported by Dirt 2 at the moment. But it really doesn't matter to me. The performance I'm getting with a single 280 is great, every maxed in game, and I even have post processing effects turned on, there's configuration option to allow post-processing ultra for non-DX11 hardware, which does seem to enhance the lighting a bit, but nothing super wowing. Now I know I'm at only 1920x1200. I don't think that 2650x1600 would be playable at these settings, but know those down a bit and you'd still have an awesome looking game at a lot more pixels to go wow at for the loss of image quality.

At any rate for my 18 month old card to still

 

[Hamish Campbell]

I have a friend who put a GTX 260 in his computer. Then he got herpes. Coincidence?

 

[Unknown-One]

I think the score was called at 10 me actually, which would be you getting owned, because you don't know what you are talking about. You have 280s, which at high resolutions are a completely different story than 260s, most especially at the resolutions we are discussing with an nfinity driver update, or 2560x1600. Maybe if you understood the concept of a minimum framerate, you could have accepted what I said as what it is, REALITY.

Prepare to insert your foot in your mouth:

My sli 260's owned the piss out of everything, at every res, except crysis.

He had GTX260 SLI, so he has had hands-on experience with the hardware in question.

If a single card cannot render a frame fast enough to display 60fps, why would three cards render the same frame fast enough to display 60fps? When each card is identical and they are not working together AT ALL to render each individual frame. Thats right, you heard it here first, AFR is called ALTERNATE FRAME RENDERING for a reason.
[...]
I noticed that in AFR modes, scenes that would struggle greatly, and guess what, per frame, alternate frame rendering makes no difference at all

Wow, you really have no clue what you're talking about. Maybe AFR mode worked like that on 3DFX cards, but its been improved since then. The drivers are set up to render ahead, which allows your two cards to work on two separate frames at the same time (thereby increasing overall rendering speed).

Now, obviously, your second card can't render a frame that hasn't happened yet, so instead they make the first card render the previous frame while the second card works on the current frame. The side effect of this is that you're always a frame or two behind what's actually happening, but as long as your framerate stays high overall, it isn't usually a problem.

Being behind by between 1 and 3 frames DOES add delay, and the lower your framerate the more delay there is. Perhaps you mistook this delay for some other form of stuttering?

Again, I'm not saying 260s are a terrible card, I'm saying they give terrible minimum framerates at resolutions over 1080p with any kind of antialiasing enabled. But the cognitive dissonance here is quite extreme.

[H]'s own benchmarks disagree. They seem to have no trouble staying over 35FPS minimum at 2560x1600 with extreme levels (16x) of AA, which means they'll perform a hell of a lot better when I run them at only 4xAA.

You also have other GTX260 SLI owners, in this every thread, agreeing that GTX260 SLI performs a hell of a lot better than you make it sound.

I So I suggest you check out some of the benchmarks online you seem so fond of, that show dual sli 280s VS tri-sli 260s. It might be an unpleasant surprise for you,too bad you never bothered to research what you are talking about, like I did in a hands on environment, and you've done only in imaginationland.

I'm not exactly sure what point you're trying to make here. Of course GTX280 tri-SLI is faster than GTX260 tri-SLI. We all already knew that. That said, the GTX260's still look like they perform great, and they absolutely OWN GTX280's on price/performance.

The reality of this portion of the thread, is you have someone with a single 260, assuming he knows what kind of minimum framerate he'll get based on what the internet says about average frames...

I've said it numerous times now, but I'll say it again since you obviously missed it. I've looked at benchmarks that compare minimum framerates, and they look good as well.

and someone with tri-sli 280s, who thinks he knows as well, and if you'd bother to look at some of the comparisons between tri-sli 260s, and even normal sli 280s, you'd know they scale quite a bit differently at high resolutions.

He said he had GTX260 SLI as well... and I did mention I looked at a comparison of GTX260 tri-SLI against GTX280 SLI, and found that they trade blows depending upon the workload.

Are you even reading the thread before you post?

I'm thinking they had SLI fermis powering the 3d multimonitor setups for a very good reason. Even the 5970 struggles to put up 60 frames on those resolutions, but yeah, I'm sure the SLI 260s will deliver a great experience at that res, and have no problem getting 60 frames. derp.

They actually had GTX285 tri-SLI powering the 3D multimonitor setups, not Fermis. I believe they were using those 22" Samsung 120Hz monitors as well, which means they were only running at 1680x1050 each (5040x1050 total).

Running stereoscopic 3D cuts your framerates almost in half, and three GTX285's were handling it acceptably. I'm running three normal 1680x1050 (5040x1050 total) monitors without any sort of 3D glasses technology killing my framerates, which should bode well for the GTX260's.

 

[sweetreddyd]

Wow, you are dense like uranium. You still don't get it. I don't think you will until you waste your own money.

'Owned the piss out of'' isn't really a quantitive figure. You think your video card owns when it's getting a framerate I laughed out of my rig.

3dfx cards rendered the scenes in sli with interleaved scanlines. Man you are clueless.

You'll notice the guy with sli 260s, uh, doesn't have 260s anymore?

I wonder why that is.

You still haven't even marginally refuted what I said. Minimum framerate in the kind of worst - case situations on gtx 260s can't keep up. You can't refute it because it's a fact. Nobody is testing games by zooming into an open trap in ghostbusters, or using vats in fallout in point lookout with multiple exploisions and transperancy effects on the screen, so they will never show you the worst possible scenario.

This is what you don't seem to understand, and I don't think will understand until you 'own' yourself by buying another midrange card that will be two generations behind at the point you buy it, expecting good triplehead performance.