The GT300/Fermi Thread - Part 2!

[Motiv]

But how many people out there have DX11 hardware?

About 1.5% if the steam hardware stats are anything to go by.

[saaya]

If what you said is true, then nVIDIA lied in a public technical specifications pdf.
And you're wrong about load balancing via the driver.
That would be the silliest move ever.
"CUDA cores" can be dynamically assigned to various tasks and load balancing should and is done by hardware not software.

Not an elegant thinking I'd say

well, when is something accelerated in hardware and when is it just being emulated? its hard to tell, and as soon as hardware detects a certain type of code or can switch into a different mode of operation which is faster for certain code, you could already call that hardware acceleration and dedicated hardware for that code... even if the same logic is actually able to process very different code... the propper definition for dedicated hardware or hardware accelration is to have some logic that is specific to a certain type of code and can ONLY process and accelerate that code and can not be used for anything else. im pretty sure thats what tesselation in gf100 is NOT... it simply would go against nvidias design goal of having a general purpose monster flop throughput processor...

so did nvidia lie? i wouldnt say so...
and even if they did, there is no way to prove them wrong, it will always be an argument of different interpretations and definitions... and in the end, like i said before... if it performs well, who cares?
will it actually perform well in a real world scenario is something that we are all curious about though... i cant wait for avp and some other actual game benchmarks with gf100...

in the end i dont think gf100 will be vastly superior in tesselation compared to rv870... it will be faster, yes, but the difference isnt 100% like nvidia first claimed... in a synthetic benchmark, the best case scenario, in a certain scene, so the best case of a best case, gf100 is double as fast as rv870... in reality the difference is probably below 50%, so 30fps vs 40fps in avp or something like that...

tesselation is NOT a killer feature for gf100 if you ask me...

[eric66]

2GB 5870 is 20% faster than 1GB 5870 in Farcry2 3X1920X1200 8XAA

Obviously 1GB vram is a bottleneck in extreme setting.

great find explains why nvidia picks psycho res and same game i wonder if we are gonna see other benchs except fc2 and unigne before fermi launchs or may be femri is designed specially with two of these in mind

[mindfury]

Asuka talked about GTX470 again,but no benchmark numbers because of NDA.
http://we.pcinlife.com/thread-1369198-1-1.html
GTX470 is slower than HD5870 in DiRT2 and STALKER COP
1920X1200 GTX470 is 10% slower than HD5870 1GB
2560X1600 GTX470 is as fast as HD5870 1GB
GTX470 TDP is much lower than 300W,but significantly hotter than HD5870

[mao5]

Is it just me or do the roof tiles on the GTX470 shot look odd? I just ran the bench and they looked much sharper and defined on my 5870.
985/1250 gave me 29.8fps, by the way.

How did GTX 4X0" kicked" 5870's A*s, DX11 Tessellation image Qualtiy Comparing

http://www.nvnews.net/vbulletin/showthread.php?t=148635

5870 kicked NVIDIA 480's A*s, DX11 Without Tessellation Performance

http://www.nvnews.net/vbulletin/showthread.php?t=148636

and this:

http://bbs.pczilla.net/attachments/m...b47f6b7056.jpg

[Marios]

about dual fermi, look at dual gt200...
http://www.xbitlabs.com/articles/vid...0_5.html#sect0
285 = 160W
295 = 200W

480 = 280W (rumored)
495 = 300W (tdp limit)
...

The official TDP is different though.
GTX295, Radeon HD 5970 and Radeon HD 4870X2 have a TDP of about 290W.
GTX280 240W, HD5870 190W, GTX285 180W, HD4890 190W full load.

[eric66]

Asuka talked about GTX470 again,but no benchmark numbers because of NDA.
http://we.pcinlife.com/thread-1369198-1-1.html
GTX470 is slower than HD5870 in DiRT2 and STALKER COP
1920X1200 GTX470 is 10% slower than HD5870 1GB
2560X1600 GTX470 is as fast as HD5870 1GB
GTX470 TDP is much lower than 300W,but significantly hotter than HD5870

conclusion charlie seems to be right if 470 is significantly hotter than 5870 i can't even think how hotter will 480 be

[mao5]

I heard NV locked the 470 speed in the Bios, nobody can oc 470 rightnow, the New AIC vendor Colorful even didn't get a 480 sample from NVIDIA.

NVIDIA hasn't decided the speed of 470

[trinibwoy]

right, ill give you that, that is pretty revolutionary... but the fact that this isnt really all that useful for games says a lot, doesnt it? as a gpu, fermi isnt really revolutionary imo and is more of a GT300... if youd use it for gpgpu then calling it GF100 makes sense... but thats just semantics

Depends on which games you're talking about. Future games will make much more use of compute shaders and hence the distinction between games and general computing will begin to fade. The software is badly lagging the hardware at this point so it's hard to see the benefits on anything more than an academic level but hopefully that changes soon.

[Funky]

How did GTX 4X0" kicked" 5870's A*s, DX11 Tessellation image Qualtiy Comparing

http://www.nvnews.net/vbulletin/showthread.php?t=148635

5870 kicked NVIDIA 480's A*s, DX11 Without Tessellation Performance

http://www.nvnews.net/vbulletin/showthread.php?t=148636

and this:

http://bbs.pczilla.net/attachments/m...b47f6b7056.jpg

I still wont reach to conclusion based on that pic. the two pics are not exactly same. There is slight difference in angle and time at which the screenshots are taken.

Will reach to conclusion when I get the 470 for the review ( have been assured of one, fingers crossed )


Little OT, just for fun :


Thats what $320 5770CF can do in Unigine 19x12 4xAA 16xAF stock speed with tessellation on

[H2O]

going above 300W is painful... you need very expensive cooling above that point, and you break compatibility with pci-e specs, so i think thats unlikely.

Saaya, I thought we had a PCIe rep confirm that neither the HD5970 nor the GTX295 were certified as PCIe compatiable, because they could pass the 300W TDP limit. So if the GTX495 goes over 300W, and Nvidia can sufficiently cool it, losing the PCIe compatiability should not be a big deal, right?

[Chumbucket843]

Exactly my thoughts. I'm not buying NVIDIA's magical tessellation performance story yet.

I have doubt that NVIDIA simply invest so much in dedicated tessellation hardware. It's probably more likely that they use the "cuda cores" to do tessellation in exchange for shader performance.

If that's the case, their DX 11 driver will have some delicated load balancing to do. It must decide how many cores should be reserved for tessellation for each game. Not neccessarily an elegant solution.


Until we see performances of real games, I'm not holding my breath.

its getting ridiculous how many people think gf100 does not have fixed function tessellation. if they are that incompetent maybe they should hire people off of tech forums to architect their gpu's. most people probably dont know the difference from fixed function logic or programmable logic anyways.

what they did was fairly simple. gf100 basically is setting up the scene in parallel compared to serial setup of other gpus. it works well for all of the small triangle tessellation creates.

[Sam_oslo]

I still wont reach to conclusion based on that pic. the two pics are not exactly same. There is slight difference in angle and time at which the screenshots are taken.

Will reach to conclusion when I get the 470 for the review ( have been assured of one, fingers crossed )


Little OT, just for fun :


Thats what $320 5770CF can do in Unigine 19x12 4xAA 16xAF stock speed with tessellation on

OMG, lets hope the angle or other stuff is making these big quality differences, because otherwise one may conclude there is a big flow in ATi's implementations of the most important future introduced by DX11. That conclusion would make the 5x-series pretty useless for Dx11 games, compared to.

But lets hope the angel, or other stuff is making the difference. Also lets hope the DX11 games never come to the marked , because otherwise he conclusion would be devastating to ATi.

[Iconyu]

its getting ridiculous how many people think gf100 does not have fixed function tessellation. if they are that incompetent maybe they should hire people off of tech forums to architect their gpu's. most people probably dont know the difference from fixed function logic or programmable logic anyways.

what they did was fairly simple. gf100 basically is setting up the scene in parallel compared to serial setup of other gpus. it works well for all of the small triangle tessellation creates.

The 5870 isn't incredibly fast on Dirt2 in DX11, playable yes but you're losing a lot of FPS. It wouldn't be that hard to prove the case eitherway, but in Dirt2 tessellation is done on top of everything else, not instead of something else, same with AvP.

I suppose in a few weeks we'll see the numbers but the fact that they are missing is worrying.

[trinibwoy]

What's there to prove? Nvidia says there are dedicated tessellation units. They designed the chip. What more do you need?

[Olivon]

[Piotrsama]

OMG, lets hope the angle or other stuff is making these big quality differences, because otherwise one may conclude there is a big flow in ATi's implementations of the most important future introduced by DX11. That conclusion would make the 5x-series pretty useless for Dx11 games, compared to.

But lets hope the angel, or other stuff is making the difference. Also lets hope the DX11 games never come to the marked , because otherwise he conclusion would be devastating to ATi.

Seems like you missed (or intentionally forgot) the images posted on page 76.
This is how Nvidia looks like, yes the same crappy quality, the trees look horrible, the roof awful.

[trinibwoy]

Enough with the humongous pics people. Olivon, what's the source for that pic?

[Sam_oslo]

[/QUOTE]

Seems like you missed (or intentionally forgot) the images posted on page 76.
This is how Nvidia looks like, yes the same crappy quality, the trees look horrible, the roof awful.


http://bbs.expreview.com/attachments/month_1003/10030521585c1936be60baceff.jpg[/img]

How could i forget such a crap? It looks very much like the 5870-results on the provides link pages. but I didn't see it, this thread runs fast you know. But you can see a HUGE quality differences with this one:



Which is provided by these guys:

How did GTX 4X0" kicked" 5870's A*s, DX11 Tessellation image Qualtiy Comparing

http://www.nvnews.net/vbulletin/showthread.php?t=148635

5870 kicked NVIDIA 480's A*s, DX11 Without Tessellation Performance

http://www.nvnews.net/vbulletin/showthread.php?t=148636

and this:

http://bbs.pczilla.net/attachments/m...b47f6b7056.jpg

Something is not adding up, but lets hope and see some more reliable sources will confirm your view, because otherwise he conclusion would be devastating to ATi, as said.

[RaV[666]]

First of all, christ people, Unigine looks pretty much identical on both ATI and NVidia, if it werent, one camp or another would call out CHEATER!
Second of all

Olivion

We have already got to the conclusion that using 2560x1600 on a 1GB card causes dramatic fps drop due to the framebuffer taking such a large amount of memory.
For the people with monstrous 30" displays there will be 2GB 5870 shortly in few different flavours(including OCed one).
So again these results arent exactly comparable.

[Olivon]

Enough with the humongous pics people. Olivon, what's the source for that pic?

http://bbs.expreview.com/viewthread.php?tid=27741

Don't know if it's legit or not

[Funky]

OMG, lets hope the angle or other stuff is making these big quality differences, because otherwise one may conclude there is a big flow in ATi's implementations of the most important future introduced by DX11. That conclusion would make the 5x-series pretty useless for Dx11 games, compared to.

But lets hope the angel, or other stuff is making the difference. Also lets hope the DX11 games never come to the marked , because otherwise he conclusion would be devastating to ATi.

did you read what mao posted and what i replied??

That thread is suggesting that 470 is compromising on image quality based on the pics posted there.

What I said is that you cant draw that conclusion as the pics are not from same angle, and so called inferior looking 470 pic may well be because of the difference in the camera angle of the pics.

Seriously, read before quoting people.

See the pic I posted, its on ATI setup and has same scene. And Trees look identical to what 470 is showing. So I was questioning the guy who reached conclusion that 470 is performing better because of compromised image quality.

[zerazax]

How could i forget such a crap? It looks very much like the 5870-results on the provides link pages. but I didn't see it, this thread runs fast you know. But you can see a HUGE quality differences with this one:

Which is provided by these guys:



Something is not adding up, but lets hope and see some more reliable sources will confirm your view, because otherwise he conclusion would be devastating to ATi, as said.[/QUOTE]

What are ATI's drivers? Some people said that Cat 10.3 boosts ATI up to 30ish fps in it

Any who, I don't get your conclusion at all, why does ATI have to worry considering that the 5870 is faster than the GTX 470 at 1920 x 1080, ties it at 2560 x 1600, and the 2GB editions are likely to beat the GTX 470 and probably get close to a GTX480....

And that's not even counting if they've bothered with a 5890

[Sam_oslo]

did you read what mao posted and what i replied??

That thread is suggesting that 470 is compromising on image quality based on the pics posted there.

What I said is that you cant draw that conclusion as the pics are not from same angle, and so called inferior looking 470 pic may well be because of the difference in the camera angle of the pics.

Seriously, read before quoting people.

See the pic I posted, its on ATI setup and has same scene. And Trees look identical to what 470 is showing. So I was questioning the guy who reached conclusion that 470 is performing better because of compromised image quality.

I have seen what you have written. Relax, no need to get jumpy.

We have a couple pictures that looks totally different quality. It deserves attention and investigation and I'm keeping the possibilities open until we see pictures from more reliable sources.

[Chumbucket843]

Seems like you missed (or intentionally forgot) the images posted on page 76.
This is how Nvidia looks like, yes the same crappy quality, the trees look horrible, the roof awful.

http://www.pcgameshardware.com/aid,6...eviews/?page=5

After extensive tests and direct comparisons we conclude that the HD 5000 AF is equal to the Nvidia driver default ("Quality” including "Trilinear Optimization”) - except some games like Crysis. With HQ-AF without "economy measures” a modern Geforce still filters more homogenous than a Radeon HD 5000.

and blurry trees on 5870?
http://img717.imageshack.us/img717/2...0517220684.jpg

the roof looks the same on 400 series as 5000 series.