The GT300/Fermi Thread - Part 2!

[annihilat0r]

480 is-going-to-be-512-sp!!!

[Your_Boss]

Oh noooes, they're site just updated.

[Funky]

Did anyone save that B3D page?
Its been taken offline.

[trueg50]

Oh noooes, they're site just updated.

Thats really quite childish, you just made a bunch of gay references and fart jokes.

[jam2k]

neliz is always right ???

What did he say ?

[FischOderAal]

Did anyone save that B3D page?
Its been taken offline.

So, what did it say?

[saaya]

Thank you saaya. The funny thing is that even if it was run on the shader core (it's not) why is that a bad thing? I'm sure if AMD did that it would be all the rage and Nvidia's fixed function hardware would be old and backward.

it reminds me of the monolithic vs mcm talk between amd and intel...

physix also was made with support from out side 3rd parties. IMO Physics has to mature still its not really ready to be used in large scale, Nvidia physix is a step who knows ATi's physics is the next one, one thing is for sure raw SP processing on ATi is hard to beat.

on one hand, nvidia at least tried, and failed...

on the other hand, id prefer no gpu physics over nvidia physix... think about it... another amd style physix would be a nightmare...
and then think of how great it would be if physix never would have existed! mhhh thats a nice thought...

Do you have a better source? Atleast neilz has a good track record.

is that even the same neliz as here on xs?
either way, fastest nvidia card in 2010 is 295... come on... i find that really hard to believe... nvidia should be able to get at least some few dual fermi cards out before the end of 2010 and beat the 295 at least by a bit...
and a super limited edition 512core fermi at nice clocks should be able to beat a 295 at least by a few percentage points...

[Behemot]

on the other hand, id prefer no gpu physics over nvidia physix... think about it... another amd style physix would be a nightmare...
and then think of how great it would be if physix never would have existed! mhhh thats a nice thought...

Bullet is Open Source, isn't it? So it's not so bad as PhysX.

I think many would agree that we need another firm like Havok or Ageia with OpenCL engine this time. And than some OpenGL games (I expect OpenCL to boost OpenGL as it is connected, isn't it?). Just like good ol' times with Half Life 2 and so

[zalbard]

I think many would agree that we need another firm like Havok or Ageia with OpenCL engine this time. And than some OpenGL games (I expect OpenCL to boost OpenGL as it is connected, isn't it?). Just like good ol' times with Half Life 2 and so

ATI needs to promote some open-source physics engine, I've been saying that for a while now...
Without any serious support none of them will come out on top and all of them will stay unpopular and underused.

[Piotrsama]

is that even the same neliz as here on xs?

Yes, he is the same person AFAIK.
Somebody can repeat what he posted on B3D?
Is what Saaya said? ("fastest nvidia card in 2010 is 295")

Thanks.

[Olivon]

Yes, he is the same person AFAIK.
Somebody can repeat what he posted on B3D?
Is what Saaya said? ("fastest nvidia card in 2010 is 295")

Thanks.

Yes, "the fastest nvidia card in 2010 will be 295" and "May-June for real availability" but this have been deleted. Don't know why ...

[annihilat0r]

HD5870 should already be on par with if not faster than the GTX 295 thanks to the newest Catalysts. I fail to understand how GTX 480 will be slower than a HD5870. First it was 60 percent faster, then 40, then 25, then 5, and now it's slower?

I think there is a common source tricking some people, including neliz and Charlie. On average, GTX 480 is going to be 25 percent faster than 5870. Let's see which informants are tricking which informees!

[Metroid]

HD5870 should already be on par with if not faster than the GTX 295 thanks to the newest Catalysts. I fail to understand how GTX 480 will be slower than a HD5870. First it was 60 percent faster, then 40, then 25, then 5, and now it's slower?

Great post. It shows how things can start 'good' and end like a 'total storm'.

Turn of events gotta love it, by the way yes I totally agree with you. It can't be slower than the HD5870. Marketing can do all sort of things

[eric66]

well no benchs except far cry 2 go figure lol and far cry 2 bench according to 5830 tests are clearly nvidia biased

[SamHughe]

LOL! I love how they gave one for ATI in those fake benchmarks.

[Sam_oslo]

LOL, that's only the calculated one. Lets hope it gives non in the real one

[Olivon]

jmke explains the thing very well in this post

[saaya]

jmke explains the thing very well in this post

did nvidia really say that though?
40-60% faster than 295? cause that doesnt make sense... at all...
a 295 is 480sps... GF100 is 512sps... how can a less than 10% increase in sps result in a 40-60% performance increase? that doesnt make any sense whatsoever...

im pretty sure they even misunderstood that nvidia statement, nvidia probably said GF100 will be 40-60% faster than the 285... thats a statement everybody heard a couple of months ago...

[Sam_oslo]

jmke explains the thing very well in this post

That's a good reading. It says that chart is based on:

"Well, after this brief review to improve NVIDIA and related improvements DirectX 11, you teach that supposedly characterize the graphs in the next benchmarking graphics from NVIDIA. L The data used are based on a review conducted by the ATI techreport.com 5970, and has been a 50% improvement in FPS on the GTX 295, NVIDIA reported as improvements in the environment of 40-60% compared this graph. ".

I don't know how accurate this numbers are, but this is really great and it is in direct contrast with those wild guesses we have been hearing.

[kklownboy]

Are you trying to say those numbers are waste of time? Those are provided by nVidia, and even tho it should hold some optimistic number, should be interesting to talk about, at least for anybody looking for news or info bout high-end GPU and Fermi.

What else are you looking for in this thread that you find the talk about the latest chart bout Fermi-performance such waste of time?

they are made up; a good translation:

The data used are based on a review conducted by the ATI techreport.com 5970, and has been a 50% improvement in FPS on the GTX 295, NVIDIA reported as improvements in the environment of 40-60% compared this graph. We could just say it, but it seems better to present the possible scenario that will move the graphics from NVIDIA if fulfills the promise of "40-60%" of improvement on their last dual graphics. That's why this is a preview, not a review, and data need not correspond to the actual card once these go on sale.

[Tenknics]

And in the past the Bulls were dominating the NBA championship and winning almost every game by 20+ pts...

haha, what? I may have missed the conversation leading up to that, but I just read that part and laughed really loud. What is that in reference to or what does it have to do with? (not arguing a point, just curious)


Not to get OT or anything but I noticed this particular thread is pretty loose at modding. Everyone is taking open shots and insults at others...Just surprised to see that with no mediating.

You guys all need to chill. Can't everyone just agree Fermi is late and is for a reason ? Its a failure, no matter how you try to justify it.

[jaredpace]

[Chumbucket843]

You guys all need to chill. Can't everyone just agree Fermi is late and is for a reason ? Its a failure, no matter how you try to justify it.

fermi is an architecture, not a product. i dont see how geforce 400 series is a failure if it hast launched yet. there are several elements of a failure that are missing. being late is the only one so far.

[Sam_oslo]

they are made up; a good translation:

The data used are based on a review conducted by the ATI techreport.com 5970, and has been a 50% improvement in FPS on the GTX 295, NVIDIA reported as improvements in the environment of 40-60% compared this graph. We could just say it, but it seems better to present the possible scenario that will move the graphics from NVIDIA if fulfills the promise of "40-60%" of improvement on their last dual graphics. That's why this is a preview, not a review, and data need not correspond to the actual card once these go on sale

Good translation, but if I understand this right, this is not "made up" or fake. This is simulated based on data/predictions provided by nVidia. Or do I misunderstand the text?

[h0bbes]

Being late is certainly a failure and nvidias PR campaign isn't very impressive either. But apart from that we dont know anything else for certain.