ATi is going to have a BIG problem if this is true, they won't have anything to counter Fermi GX2 even if they make a 5990 or something like that...
Printable View
ATi is going to have a BIG problem if this is true, they won't have anything to counter Fermi GX2 even if they make a 5990 or something like that...
I'd love for it to be legit. But something is nagging at me, telling me that if it was. NV would have shipped some cards out by now for some preview results at sites.
On the other hand even if it's the same as ATI cards or slightly better it's something I might buy depending on pwoer usage and size.
As another poster has already mentioned, just because the ATI results seem to be realistic, it doesnt mean that the fermi results are true as well. Wouldn't you expect the person that created this chart (assuming its fake), to at least use some of his brain and put down the correct numbers for the ATI cards?
All this does is build peoples' hopes so high only to get crushed into a thousand pieces when Fermi is released.
Of course that would be a possibility. But if that is the case, the user certainly did a good job mimicking the exact layout Nvidia uses for their official PR slides. It could really go either way, but my personal opinion would be that there is more evidence suggesting it is real than not.
What I'd really like to know is more information on the user who originally posted these and what their connection with Nvidia is. Perhaps NDA has been breached?
So you have access to the real benchmarks, I see?
I just wonder about the prices. GeForce cards were always more expensive than equivalent Radeons, so if HD5890 costs 500 eur, GTX 360 will cost like 700 eur...
I'd love for it to be real as well. Unfortunately it's not. Although it's a new architecture, it was built around GPGPU. This isn't like the G80. I just don't see the performance jump in gaming to be this big. If it were, Nvidia would be boasting about it and not just its GPGPU performance which we already know is going to be great.
There's no connection. It was posted by a troll. Read page 4.
Anyone can make a slide like nvidia if given enough time, which this poster had.
Also, judging by your posts on this board and others, you said yourself Nvidia is not going to make the same mistake in pricing as last gen. With these kinds of performance numbers, I'd find it hard to believe Nvidia would price their cards that way. Heck, with these numbers they could price it at $700, which you said yourself was not going to happen.
Exactly alike? Lol. Go ahead and find some nvidia slides for us that look like that.
Are you suggesting that these results are similar to those of your "NDA source"?Quote:
What I'd really like to know is more information on the user who originally posted these and what their connection with Nvidia is. Perhaps NDA has been breached?
Believing these slides is most likely going to let you down when the real cards come out. If you had more realistic expectations for the card then you probably won't be let down by the actual results.Quote:
So you have access to the real benchmarks, I see?
nVidia's trick to stop people buying AMD?
It's just nvidia's way of putting off ppl buying ATI for upcoming holidays. :D
Doesn't matter fake or legit or trick,
the real thing we know it's faster than 5870, but how much faster
The only concern me is the price , release the dam price chart would you nvidia
*And then crickets from 003's corner.*
Here is a hint. If you had an NDA source and have seen the performance as you'd claimed then you wouldn't have to wonder if it was an NDA breach. You could tell us straight up if these numbers are real without putting your source at risk.
And it's too big to fit in the trash can, lol.
Guru3D Hilbert Hagedoorn obviously knew the slides are fake so why does he use a forum post by NVIDIA fanboy as news, obviously getting the hits is more important to him than credible news.
Maybe somebody should post at GURU3D forum some ATI HD 5890 slides with awesome performance beating GTX380 and see if Hilbert will use it as news.
The reason is, I believe the ATI fanboys deserve at least the same excitement as the NVIDIA fanboys are getting from the NVIDIA fake slides in this thread, so why not post some fakes for them.
If you only could... supply here in europe is anemic... and cards sell for 400-500€ :p:
I was lucky that i scored my Powercolor LCS for only 420€... price gouging right now is insane... you pay the same price i did, for a plain vanilla HD5870.
If supply is still that low next year i might can sell my card for more then i bought it. :ROTF:
get ready for ATI optimized drivers!
1920x1200 could be cpu limited in Nvidia's benchmarks. The 5970 does not really shine till you give it 2560x1600 to play with. If the GTX380 could bloe the 5970 away at 2560x1600 then I was be impressed. If it only ties the 5970 why wait 6 months for something you could have now?
+1 we know it will be pretty much 99% but it will sell on price point.. its its 50% more expensive and only 25% faster then.. well only the hardcore will buy it...
and even then I still think ati have a LOT of room to wiggle on price
its just suspicious seeming its NOT an nvidia slide (is obv comparing them to their others/old slide) and also its posted by a guy called "some troll" iirc :/
Note, the benchmarks are Nvidia friendly games. Even if these are accurate benchmarks, I doubt the performance would be so good across ALL games.
Also, its not so off the wall for GF100 benchmarks at this point. Its not more than a few days ago that Nvidia's Tom Petersen was showing off his GF100 SLI rig, so Nvidia def has some working cards.
But like I said, its all hearsay at this point. No reason to get wound up until the cards are finally released (at least a paper launch).
One thing I will say, these better not be too far from the truth, or Nvidia will be faced with a R600 of their own...
Nobody benchmarked blockbuster games like Batman: Arkham Asylum with innovative features like PhysX® technology?
Umm, holy crap. Does that mean the GTX380 will be priced at ~$600?
Don't make me bring out Sam's balls benchmark graphics again. Sick and tired of this useless Fermi threads with fake graphs and info.
Well, I think it is fake.
http://www.guru3d.com/imageview.php?image=21418
Stalker CS, maximum settings, 1920x1200, and a GTX295 with 63 fps? HD5870 57 fps?
back to reallity
http://www.xbitlabs.com/images/video...r/07_stalk.png
It could be a different benchmark run. If I look at the sky the whole time my framerate looks much better.
On the different hardware forums it's funny how many Nvidia users have started listing their cards as of today :lol:
If the 40% faster then Cypress claim is true then the numbers for GTX380 would be 132, 80, and 80 for RE5, S:CS, and FC2 respectively. Impressive, but still slightly behind the 5970.
i would love to see fermi own ATI, nothing is wrong with jumps in performance. i dont care which company does it, cause in the end, it means better prices per frame for us.
however guys who think its not possible to get a good jump just on architecture, 3879 to 4870 was quite shocking and did just that. however i will also say i dont believe these benchmarks. just looking at the really bad xfire scaling, its looking to be way to much marketing BS and not enough truth. and honestly if it is accurate, then once ATI gets the xfire drivers good and we see 80% scaling instead of the 30% shown, then it puts 380 right between 5870 and 5970, which is what we all expected. (basically if they removed 5970 from the charts, would any of us really be surprised by the numbers?)
LOL yall 6 pages in hours ...& the degree of careing is astounding !
Heres to compitition :clap: and priceing :clap: because in a year or two I will get a fermi card from one of you for a great price :D
THIS :yepp:
maybe you were only focusing on that point of the post rather than what the whole post meant. being that you quoted just that sentence.
It means if true, 5000 series will drop heavily in price second hand, making them cheap "monster" performing cards. which is obviously a good thing.
One could only dream this was true. Imagine getting GTX295 performance for under $300, or even $200 depending on how aggressive Nvidia prices their cards? It would be madness! And a great benefit to EVERYONE as it will truly open up PC gaming to much MUCH more people.
Then again.. sadly, this is a fake slide :(
obviously, waterblock. Duhhhh;)
lol on the second part.
the 5870s fan profile is the problem.
You guys know you can adjust the default fan speed on the the stock coolers right? I use MSI afterburner's default profile manual control profile The gpus at the 5970 idle around 34C and load they not go beyond 70C. I even ran my 285s at above stock fan speed. Under load they required 100%, the the gpu temps on those were 80C-85C. It's too bad Nvidia likes to stick with crappy fans on their coolers. I'm not even going to talk about them still using aluminum :shakes:
Why are you skeptical of its legitimacy? There is nothing to be a skeptic about. There is nothing about these slides that could cause even the slightest doubt that they are fake. They are simply a more official-looking remake of the 1920x1200 data in the THG graphs (well the Far Cry one is) which as I have said more than once on several forums are completely fake as well.
i think every benchmark on the new cards are rubbish anyway even if they are true. for me i think its all down to how they perform in DX11 especially as i will be upgrading when all new games will be DX11. wait till DX11 beenchmarks are out then we'll see who really has the monster cards. end of thread!:cool:
successful_troll is laughing his ass off
well this may be fake but i would expect that kind of numbers from fermi, i mean what less for 512 shaders , thats more than 2x GT285
5970 and 5870 numbers dont seem genuine, c'mon GTX360 matching 5970 is blasphemy :p:
Look what Cartman crapped...
http://i46.tinypic.com/2qdoqbd.jpg
Nice one:rofl::ROTF:
I can doo more..
Games aren't shader bound so the shader increase isn't enough to predict performance. Until we know more about the architecture (and clock speeds) it's impossible to know how it will scale. Cypress didn't change much at all architecturally over RV770 and it doubled everything except bandwidth and only managed 50% more performance on average.
please 'doo':D
Here is one more for the road...
Larrabee unleashed to the public.
http://i50.tinypic.com/sfar2c.jpg
In this post he was refering to the FPS AMD's cards had. But in another post he says this:
http://www.abload.de/img/lebowskiyeahryyh.jpg
Tom put on a few kgs heh
really nice guy i hope those figures are real and cards come out asap we need some competition or AMD prices will start going stupid like Intel CPU prices are right now
This is the post I was referencing to. And when I said it "looks legit", I was referring to the fact that the FC2 performance figures on the 5970 were in fact accurate in real life, and that the graphs do appear as they are from Nvidia.
In fact, saying they look legitimate is still a perfectly acceptable statement, even now, because they do in fact look legitimate. Whether or not they actually are is a different story.
Also does anybody have a link back to the original thread where "successful_troll" posted these graphs? Because I did see some graphs posted on Tom's Hardware by a user named "successful_troll", and these graphs are completely different from the ones I saw.
This is were these slides appeared the first time, no source, nothing.
http://forums.overclockers.co.uk/sho...php?t=18089093
Why isn't the name of the poster "successful_troll" then? Or did people just get that mixed up with this?
check the results. THey are identical to the ones from the THG forum. They just took the numbers and put them in the slides. If there were from different sources, i doubt you would get identical results for all the cards.
Well i am thinking that successful_troll (who posted the benchmarks in the tomshardware forums) is the same person as svs22 (who posted the benchmarks in the overclockers forums). This might be stupid reasoning but:
From tomshardware thread:
successful_troll
This benchmarks is not fake, is from the tomshardware review about the GTX 300 series.
Looking at those benchmarks the GTX 3xx seems to be more powerful than we expected, I am surprised.
From overclockers thread:
svs22
Looking at those benchmarks the GTX 3xx seems to be more powerful than we expected, I am surprised.
this is it i think:shrug:
http://www.tomshardware.co.uk/forum/...6100_15_0.html
http://www.tomshardware.co.uk/forum/...6100_15_0.htmlQuote:
successful_troll
This benchmarks is not fake, is from the tomshardware review about the GTX 300 series.
Looking at those benchmarks the GTX 3xx seems to be more powerful than we expected, I am surprised.
it is the 8th post on the page. Now 003 please stop trolling
:Quote:
successful_troll wrote :
This benchmarks is not fake, is from the tomshardware review about the GTX 300 series.
Looking at those benchmarks the GTX 3xx seems to be more powerful than we expected, I am surprised.
Quote:
MARSOC_Operator wrote :
Care to post a link then? I think you fail.
:shocked: omg he's telling the truth!...Quote:
...404 error figures:down::shakes:
so its still a mystery to the gt300's performance as there was clearly no review, and there wouldn't be one anyway since nvidia havent even made a benchmark yet. i cant believe one guy made so much hype.
haha, i thought you already knew the link was fake...
But forget about Fermi, we must wait for the GPU to rule them all, Larrabee... :rofl:
http://img694.imageshack.us/img694/8756/locked1.gif
It's probably not a good idea to argue constantly and call people trolls when you just got your foot in the door.
I don't want you to think I'm taking sides, so click on your user name and find more post by "your name" and look back through them.
Do the same for the person you accuse, then ask yourself who needs the...http://img705.imageshack.us/img705/9302/trollspray1.gif
.
.
According to this post, the slides are fake per nvida.