RV770 really turned it around for ATI, especially after you consider the debacle that was R600 vs G80
Printable View
RV770 really turned it around for ATI, especially after you consider the debacle that was R600 vs G80
I think at the very worst, we are looking at a 2900xt like launch. And even then it will be likely better than that launch because even Charlie thinks this card will be faster than the 5870(while the 2900xt wasn't faster than the gtx8800, used more power and it was six months late). Even if the scores are indeed 5 percent better, with better drivers, I think we can squeeze out another 10 percent or more since this is a totally new generation. Heck 10% is really conservative since this is totally new technology. If Nv gets to the source of what is causing the poor clocks and implements this into the refresh(still at 40nm), and gets up to its target clocks, we should be getting a card by the time the gtx 485 comes out that is indeed 50% faster than a 5870 even by Charlie's initial estimates.
However if the 1200mhz clocks are true, it might look like the gtx 480 is a 2900xt just dying for a refresh.
don’t know if it was mentioned before but here it is just in case
Hitler reacts to Nvidia Fermi Benchmarks
Wow! Just Wow! Another month before the official launch and then probably a month or two for the cards to be on the shelf at MSRP price, so then a month or two more and the price then becomes competetive and then we have the HD6000 launch. Not looking good for Nvidia. Not looking good.
As far as I remember, we have always had GPU on the shelfs on launch-day in the resent past. Where did you hear it would probably be a month or two for the cards to be on the shelf at MSRP price?Quote:
Originally Posted by Utnorris
Demand and supply, it's as easy as that Sam_oslo :) Our (maybe not) allmighty Jensen already said that it will take some time until Fermi is available in volume (and if the reports are true, it will sport A3 as well).
While at first cards might be near MSRP, the price is likely to increase as soon as the shops realise that. Remember the launch of the 8800 GT? :) A few were lucky and got the card early and comparably cheap, the people who waited had to pay more.
The exact same happened just a few weeks ago with the HD 5850 and HD 5870. NVIDIA/AMD set the MSRP but in the end it's nothing more than an advice.
Hot off the Twitter:
http://i56.photobucket.com/albums/g1...n/e19e8ef0.jpg
They are just trying to save their butts IMO.......Paper launch for sure or super scarce supply coming at the end of march so they can say they "launched it."Quote:
Originally Posted by ShadowFox19
well on official site they're not mentioning word launch:Quote:
Originally Posted by 96redformula7
http://www.nvidia.com/object/paxeast.html
this thing will be so scarce that you'll need to a) seen it at PAX b) preorder it... 'cos you'll be only able to "kick yourself in the next few months" :DQuote:
Originally Posted by NVIDIA
i dont think its their marketing department, its their sales department...Quote:
Originally Posted by damha
i guess cause they know they wont have GF100 parts for a long time, they will probably only have their initial 10K wafers of GF100 and the real availability they mentioned for Q2 will probably be with GF104...Quote:
Originally Posted by zalbard
hmmm is it only me who finds it a bit ironic to bash charlie as unreliable and then mention almost the same points he did? the only difference is that he doesnt believe it will hurt nvidia that much. :shrug:Quote:
Originally Posted by onethreehill
oh nvm, so it was not just me :DQuote:
Originally Posted by SKYMTL
pff right, like ati would be that stupid :lol:Quote:
Originally Posted by tajoh111
hahahaha whaaaaaat?Quote:
Originally Posted by Sam_oslo
nvidia has been milking the market with G80 derivatives for how many years? and now after ati has the performance crown for what, 4 months, and sells some highend cards for more than 500$, they have been milking the market for too long? :rofl:
there definately wont be cheap and good nvidia gpus "flying around everywhere" for quite a while... wake up man!Quote:
Originally Posted by Sam_oslo
he didnt say that, did he?Quote:
Originally Posted by FischOderAal
he said it wont be notably faster than 5850/5870, and then ads that in his opinion nvidia doesnt have to beat them anyways... :D
whenever somebody talks about why he doesnt have to do something, hes only trying to disctract from the fact that he CANT do it :D
i dont think charlie cares all that much about hits man... :shrug: hes writing the way he thinks and feels about nvidia, easy as that... hes not hiding his anger towards them, nor is he making it up to get more hits with drama...Quote:
Originally Posted by SKYMTL
it doesnt say launch anywhere...Quote:
Originally Posted by xdan
so where does it say launch?Quote:
Originally Posted by xdan
all this says is that there will be at least one fermi based system you will be able to play games on... plus a speculation that maaaaybe there will be pre-orders at the event...
thats not a launch... if they would launch it then they would mention it big time.Quote:
dont miss the launch of our next gen gaming card!!!! it will revolutionize the gaming experience, a once in a lifetime event!
if the 5830 is priced right, cypress might actually turn into a 8800-ish card :DQuote:
Originally Posted by ***Deimos***
but tbh... its impossible for any card to follow in 8800s footsteps now cause there simply arent any demanding games out right now that would require an upgrade... :shrug:
:D :up:Quote:
Originally Posted by BenchZowner
no, it wont be that bad, nv30 was SLOWER than its predecessor in quite some games! fermi cant possibly be slower than a 285 seing as they use very similar building blocks, fermi has more mem bw and even the 470 has double the sp's as a 285...Quote:
Originally Posted by kadozer
yeah, thats more like it... and on ln2 fermi will probably spread its wings if its that heat limited as it seems to be, and itll have the benching perf crown, even if it cant be bought anywhere or sucks on air cooling...Quote:
Originally Posted by tajoh111
9k wafers total of the A silicon.Quote:
Originally Posted by saaya
Not all those wafers will be ready for launch, they have to space out the availability until B1/B2 is ready, late Q2/early Q3 if the B silicon rumor is true.
:rofl::rofl::rofl:Quote:
Originally Posted by firas
I like these fermi threads, so much fun :yepp:
super saaya super quote. That is all.
I think you are given to much credit to Charlie for his intentions. If he was a journalist he would write like one.
Charlie wants to get as many hits as possible. This is a man who will dress in a bunny suit at a convention. Who will hand out pamplets which try to humiliate the company that is having a function there. Charlie tries to write something that will provoke as much of a reaction as possible. Charlie isn't conservative in his writing because it helps get hits. It makes him look wrong alot of the time, but in this day and age, you don't get hits with vague uninteresting information. You do it by going out a limb and saying hard numbers(even those it might make you look like a fool in the end).
Charlie is trying his best to get people to notice him and is an attention whore. It probably why he started his own website. He will do something that demeans him just so he can get more people to read his stuff or listen to him.
Now i know why the multi-quote button changes to Saaya :DQuote:
Originally Posted by [XC] gomeler
Anyway i hope Fermi perfoms better than NVs PR team does lately. We need competition to keep prices in check.
The problem is.. he has not been that far off. And unfortunately... we have been lied to by "reviewers" and BIG QUOTES "Internet journalists" for so long now (hey remember when nVidia PAID people to go spam forums??? was not that long ago ;) )ANYTHING you read should be taken with a grain of salt because 99.99% of them NEVER went to journalism school and as such never took an ethics class. On top of that.. you gotta figure that 9 out of 10 journalists are slanted anyway.Quote:
Originally Posted by tajoh111
So what did Charlie say wrong? 2% ? TSMC admitted to "single digit success percentages" on their wafers but that the yields were only as low as 20%. (If you cant find the story on google you either are not looking or it was removed) March release date? Yup.. looking like it will be a paper launch at that time... Poor Clock management and overheating problems? Why is it that when he said it 5 months ago it was terrible slanted news and now that it is looking to be fact.. he is still be lambasted? I dont respect the guy... his writing is terrible, but he is right. (for a blogger) At least he has been pretty spot on with Fermi. The fact that they are releasing NOT their flagship model in march... (which is utterly unlike nVidia) should be s ign that the guy was not wrong.
He maybe he is (and I am) and there will be more than 1000 release units of the 480 in March... I expect it to hit stores in May though as he predicts, because lets face it... he has been right thus far...
So Charlie was right! but NOT, in the same breath. And the fools that think Charlie is an idiot but take nVidia PR as gospel (a company that has been proven to lie to you over and over) well I think that some "doth protest to much."
Trashing Charlie is kind of like someone joining a discussion about network news and mentioning the daily show. (Even with the caveat.. "I dont think it s a news show but I am just saying..."). Thanks a lot man.. .way to bring down the intelligence level of a discussion. Talk about a comedy show and try and compare it to news... especially since it IS completely slanted... well whatever... stupid point about stupid people.
So with that thought process... Don't believe ANYTHING you read.. but also don't try and discredit Charlie either... he is more believable than the great green PR machine.
No one, especially this board has been taking NV pr as the truth. Similarly no one has paying attention to the pr of other company. I don't think NV should be singled out for lying, all companies do it nowadays. I don't remember Nvidia even really lying about fermi. Its just rumors that didn't turn out to be true more than anything. Sure rebranding g92 can sort of been seen as lying, but they are not forcing a gun to anyone head to buy it and they are not raising the price. The only clear lie I think was the fake gf100, but that wasn't going to do anything for potential buyers.Quote:
Originally Posted by Neuromancer
Although, Charlie has been right a few times(all related to the same thing), doesn't make him the most accurate guy. Your making it sounds like charlie hasn't been wrong before. Charlie has been wrong on a few things, the most recent being NV cancels big bang MSI motherboard or oakridge project being canceled. Or Nvidia being in trouble for the last two quarters for having nothing to counter 5xxx.
I don't remember about hearing about people spaming on behalf of NV. I don't think you are giving review website enough credit if you think 90 percent are being bribed. On the opposite end, people say anandtech was one of the people who brought about the gts250 coverup, but these same people will then say Intel is known for loving intel and from what I have seen, only Anand has a bawlin lifestyle supported primarily through his website. Similarly, it was not that long ago when HardOCP was being accused by members on this board for favoring Nvidia. I think there is alot of conspiracy theories that are not true.
I thinks it dont buy things based on rumors and to also buy something when you actually need it. Don't be a fanboy for any company because all they want is your money.
Too much I think and I don't think in this post. You are missing much history.Quote:
Originally Posted by tajoh111
The one who raises the biggest stink gets the most attention. No PR team is honest, but it is all relative ;)
Ok I think you are missing my point... because I am not good at elucidating no doubt.Quote:
Originally Posted by tajoh111
First of all I see people proclaiming the best of Fermi since November. When the product did not even exist.
I also am not claiming that Charlie is a font of truth. I am just a little sick of seeing greenie fanboys trash him every chance they get, why him and not every other stupid blogger in the world? It is starting back up again because fermi is "possibly" being released next month. The thing that bothers me most... is the guy is pretty intelligent and seems to have some good information and is more right than not, which is pretty rare in the forecasting world. Especially knowing hte strangle hold nVidia has on journalists.. so going against them it is REALLY hard to get truthful information from sources. I am sure he has gotten bad info as well. Let's hope he some shred of journalistic integrity, and actually validates from more than one credible source.
Is he always right? No, has been right with Fermi? Seems to be true... I dont work at TSMC but know someone who did, and heard about the 45-40nm "shrink" (this affects ati and nvidia) both are built on 45nm process basically... its out of my ken to discuss ...
Yup. PR is PR.. does not matter if its a company trying to sell a product, or a blogger trying to make a name for himself. EVERYONE knows who Charlie is now...well every geek anyway. It does not matter that everyone at XS knows that that guy is right or wrong 100% of the time.. 99% of the world is taking it as news... that is bad.. because most people do not know how to separate the wheat from the chaff.. he has a lot of wheat.. covered with a lot of chaff.Quote:
Originally Posted by damha
GPU politics is just as bad as any real world politics ....
luckily it just 2 GPU manufacturer for now
Do you guys forget the name of Charlie's site? Think about it..it would save you all a lot of time.
I highly doubt we will see very low availability of GF100 cards when they do hit the market like we did with with ATi's 5xxx cards during the past 6 months since It's quite clear the setback was caused by poor yields in TSMC's .40nm production. ATi rushed out their cards on this same process which caused them to have very high production costs and near non-existent availability during 2009 with the hopes that they'd grab a sizable chunk of market share away from nVidia. So far, this tactic has not proven itself to be a failure when you look at quarterly sales figures and dividend returns. To my surprise, a large majority of customers who I build high end systems for have yet to show any inclination in going in the direction of transitioning to HD 5xxx graphics solutions even after I inform them that ATi cards are clearly the best performing cards currently on the market. Many of them simply don't want to go with ATi because the lack of Cuda and PhysX processing, a history of shoddy drivers, and because they think ATi's current market lead will be just like previous leads in that it's typically followed by a couple years of being second rate once nVidia gets it's new architecture on the market.
A lot of consumers see ATi's performance lead not as some huge feat of technological innovation, but rather as the result of nVidia holding off release of a clearly better card because manufacturing costs would have been extremely high (as ATi has found out) due to TSMC's very low .40nm yields last year.
If ATi's strategy was as good as the fanboys here claim it to be, why has it clearly failed in terms of profits and market share gains? The goal here for ATi was to release a faster card and claim the performance crown while propagating the idea that nVidia's upcoming card had major flaws and will be another NV30.... all while hoping consumers will turn away from NV. And even though their insanely vocal fan base followed this strategy perfectly... the message has not has achieved any of it's goals as the numbers below clearly show.
I personally believe many consumers are completely turned off on buying ATi products when they go to websites like XS to research a pending graphics card purchase because of flood of negative crap consistently flooding any thread where readers simply want to get accurate, fair, and current information.
http://www.kalionzes.net/computers/amd_nvda.jpg
anyone want to guess what will be the next fermi joke ????
http://www.xtremesystems.org/forums/...3&postcount=70
Quote:
Originally Posted by Andrew LB
Wow best post in 6 months and of the whole thread. QFT.Quote:
Originally Posted by Andrew LB
It's not an instantaneous effect, it takes time for an advantage to impact market share. However, 6 months is nothing in the grand scheme of things and it always amuses me when people get all hot and bothered over short term swings in momentum. AMD has to keep this going for quite a while longer to make a big dent in the market.Quote:
Originally Posted by Andrew LB
i see a GTX480 ultra rare edition with some fancy cooler retailing for 1200$ on ebay soon ..... maybe it will even have a unicorn sticker on?????
lol talk about foot in mouth
He also claimed the high end Fermi part had been castrated to 448 SPs.Quote:
Originally Posted by tajoh111
^ How do we know that's not the case? I haven't seen any reliable sources stating the GTX 480 will have 512 SPs.
Yes 512 sps has only been stated by Fudzilla.
GF100 White paper on nVidia site, page 12Quote:
Originally Posted by trinibwoy
Quote:
Originally Posted by trinibwoy
Wrong. Nvidia explicitly unveiled the GF100 architecture to have 512 SPs at the Deep Dive event.Quote:
Originally Posted by kadozer
:up:Quote:
Originally Posted by Olivon
Quote:
Originally Posted by Olivon
Perhaps you need to read my question again. I didn't ask about the GF100 architecture. I asked about a specific product based on it.Quote:
Originally Posted by ElSel10
Charlie himself said top GF would have 512sp in his latest article.Quote:
Originally Posted by trinibwoy
even charly(tan) says so! so it's true. :p:
Sigh, why is nobody reading my question properly? I said reliable source :p: Charlie also claimed it would have 448 in an earlier article, he's carpet bombing with his predictions as usual so he can claim to have been "right" once the dust settles.Quote:
Originally Posted by annihilat0r
And what is a reliable source ? B3D forum or what ?Quote:
Originally Posted by trinibwoy
Please be serious ...
.Quote:
Originally Posted by Andrew LB
.
.
.
.
.Quote:
Originally Posted by Andrew LB
.
.
.
.
.
I laugh so hard, i barely avoiding pissing in my pant. :ROTF: :rofl:
Way to go fella, keep on making prediction & "smart" post like those. :clap:
two nvidia boys agreeing on same bs never saw that coming :ROTF:Quote:
Originally Posted by NaMcO
It can't be Nvidia PR :ROTF:Quote:
Originally Posted by Olivon
not a reliable sources.Quote:
Originally Posted by Olivon
also not reliable , but if he said it then it must be true :DQuote:
Originally Posted by annihilat0r
just now i took a quick look at the semiaccurate forums, man they hate XS so much :D, funny how Charlie can't be accurate about the number of people who's stalking him, he says 3-4.
Nice quote there :DQuote:
Originally Posted by mindfury
Having your cards/architecture ready is "rushing" in your book?Quote:
Originally Posted by Andrew LB
You might want to see Nvidia's struggle to put their fermis on the market and rethink the whole situation. Because, they are the ones rushing their launch to march, when real availability is slated for may-july.
Somehow (based on your posts lately) I would rather believe it's actually the reverse situation, your clients ask for the new Direct X 11 cards which they see are faster and everybody is talking about, and as you don't like ATI.... just give them those so important reasons as "lack of Cuda".Quote:
To my surprise, a large majority of customers who I build high end systems for have yet to show any inclination in going in the direction of transitioning to HD 5xxx graphics solutions even after I inform them that ATi cards are clearly the best performing cards currently on the market.
Many of them simply don't want to go with ATi because the lack of Cuda and PhysX processing, a history of shoddy drivers, and because they think ATi's current market lead will be just like previous leads in that it's typically followed by a couple years of being second rate once nVidia gets it's new architecture on the market.
And about the drivers.... I've used Nvidia since 2001, but for my last purchase I chose ATI (my first ATI), and after 1+ year using this card....I had zero problems, and games played perfectly.
Because of the same problem you just mentioned: TSMC process/yields problems.Quote:
A lot of consumers see ATi's performance lead not as some huge feat of technological innovation, but rather as the result of nVidia holding off release of a clearly better card because manufacturing costs would have been extremely high (as ATi has found out) due to TSMC's very low .40nm yields last year.
If ATi's strategy was as good as the fanboys here claim it to be, why has it clearly failed in terms of profits and market share gains?
I didn't get the whole picture in all that twitter-mess last might. Is the official launch date for GTX 480 & 470, March 26 now?
3dmark Vantage Extreme and Performance on GTX470 (no screenshots):
http://translate.google.it/translate...n&hl=&ie=UTF-8
X73xx and P167xx (Physix?)
for reference:
http://www.tweaktown.com/reviews/314...rd/index4.html
yep :yepp: of course nvidia failed to deliver message that simple instead pissed of ppl with that lan party anouncement :ROTF:Quote:
Originally Posted by Sam_oslo
So this puts GTX 470 to around 5850-5870 level performance?Quote:
Originally Posted by Gilgamesh
Frankly, I hope no.Quote:
Originally Posted by annihilat0r
Bad/old drivers?
I've done some research and concluded that those Performance scores are in line with HD5870 and Extreme scores are a bit lower than HD5870.
But obviously those HD5000 scores are from initial Cypress reviews and I believe those scores should have gone up slightly with newer drivers.
However, the same should be the case with GF100, it should increase performance over time. Even more so than Cypress, taking into account the new architecture.
If the 3D mark vantage scores are correct this is bad. I mean in extrema GTX 470 actually loses out to 5850 Toxic "765 clock instead of 725"
But i also hope its a driver thing, because if its not a mild boosted 5870 would be able to compete against a GTX 470 :(
I would prefer 19xxx and 9xxx from GTX 4xx !!
@annihilat0r
new cat.10.3beta give 1000+ point in P preset of Vantage on HD5870, so you're right.
Besides we don't know if these scores are PhysiX on or off ...
I'm not so sure about the those Tweaktown scores. I mean, other scores around the Internet give HD5870 around 16500P and 8000X. But maybe that article was written with better drivers or a better configuration.Quote:
Originally Posted by ajaidev
Wow, that's amazing. And I still haven't installed those drivers! Such powerful a card is HD5870, I didn't even need to overclock it whatsoever or need better performance at all. No PC exclusive games = No need for better performance.Quote:
Originally Posted by Lightman
I don't know how an entire forum can be considered a source but I'm not sure what you're asking. You do know what reliable means right? Trustworthy? Dependable?Quote:
Originally Posted by Olivon
The only SKU announced so far has 448 shaders. All the talk about 512 shader parts being available in the first batch of Geforce cards has been from forum posts and rumour mongering sites - i.e. unreliable sources.
"GeforceGTX470 in somehow in the 3Dmark Vantage is divided into sub-X7400 around, GPU score of 7100 points, while the same platform Overall score for 6400 run RadeonHD5850 around, GPU score of 6200 or so. 同时据称GeforceGTX470的核心/显存运行频率为810/3200MHz，采用320Bit GDDR5显存。 At the same time allegedly GeforceGTX470 core / memory operating frequency of 810/3200MHz, using 320Bit GDDR5 memory."
http://translate.google.co.in/transl...en-US:official
ahhahhaa 810/3200MHz wazza ??
810MHz is impossible,even GTX480 core clock will not pass 750Mhz....
PS:
Core clock in GF100 is GPC clock（half of hot clock），not L2/ROP clock...
Lol what, GTX 470 is 810 MHz? No way!
And if those scores are really from a 810MHz 470, then all the worse!
LOL, talk about getting your expectations high.Quote:
Originally Posted by ajaidev
Since both scores are around 7k lets give some slack to nvidia and say its x7400 overall. Now my friends 4870 x2 gives about x7350 and my GTX 275 OCed gives about x5500, 5850's with better drivers such as 10.3 beta may give x7200 and 5850 overclocked may be able to give x7300.
So overall overclock 5850 would give about the same score as the proposed GTX 470 score? This is very troubling indeed also i am about sure that the 810 Mhz is wrong.
From when 3Dmark vantage scores were important ?
I want to see a real gaming benchmark
Xbit Labs is reliable or not ?Quote:
Originally Posted by trinibwoy
Quote:
The flagship Fermi graphics processor will feature 512 stream processing engines (which are organized as 16 streaming multi-processors with 32 cores in each) that support a type of multi-threading technology to maximize utilization of cores. Each stream processor has a fully pipelined integer arithmetic logic unit (ALU) and floating point unit (FPU). The top-of-the-range chip contains 3 billion of transistors, features 384-bit memory GDDR5 memory controller with ECC and features rather unprecedented 768KB unified level-two cache as well as rather complex cache hierarchy in general
Flagship doesn't necessarily mean GTX 480. For what we know the yields of the next revision might be good enough to harvest quite a few fully functional chips with 512 CUDA-cores. These chips then might be used for a Ultra 490.
Pure speculation on my part.
Additionally, the part of the report is just a list of what is known from the whitepaper... Xbit Labs doesn't claim GTX 480 will have 512 CUDA-cores either.
They are just regurgitating high level specs from the whitepaper. I'm talking about the specific configuration of the GTX 480 SKU.Quote:
Originally Posted by Olivon
GTX 480 has 512SP, i thought it is a taken as a facts, where are you was last few weeks/months triniboy?
I'd say not as their story says it's 384-bit, it is almost certainly 448.Quote:
Originally Posted by Olivon
So overclocking from 725MHz core to 1GHz is only 100pts? I think not.Quote:
Originally Posted by ajaidev
According to who? We have a million rumours about everything including the color of Jensen's underwear but hardly any real info on launch configurations, performance or pricing. We all know the chip has 512 SPs but that doesn't guarantee they will launch a card with all SM's enabled on March 26th.Quote:
Originally Posted by Zed_X
Are you joking? First of all, the codename GF100 explicitly refers to the GeForce line of cards. And the Deep Dive was dedicated to the top GeForce part as well. And in the whitepaper, it says it refers to the top GF100 part.Quote:
Originally Posted by trinibwoy
What are you suggesting, that nvidia lied in the whitepaper and at Deep Dive? :rofl:
Now you're just contradicting yourself and from the looks of it, making an attempt to troll. I thought the whole point you were asking was because we didn't know the chip has 512 SPs? And going from Nvidia's last generation, the GTX*80 was the top end single GPU part. There is no reason for that to be different this time around and we know the 480 and 470 are launching at the same time.Quote:
Originally Posted by trinibwoy
And when even Charlie has said it will have 512 SPs that's when you know it's time to stop arguing otherwise.
The 480 will have 512 SPs. There is no need for a discussion about this fact.
The education system is really failing humanity. GF100 is a chip. GTX 480 is a part based on that chip that may or may not have all SM's enabled. You know how the top announced Tesla SKU only has 14 of the SM's enabled though it's based on the same chip? Why is such a simple thing so difficult for you to understand?Quote:
Originally Posted by ElSel10
@mapel, it's only a fact when Nvidia announces the specs. So far we only have random people on forums asserting "facts".
http://www.nvidia.com/object/gf100.htmlQuote:
Originally Posted by trinibwoy
Look in both PDFs. 512 Cuda Cores everywhere.
I am following the GPU-Buiness since 10 years. There was never the case, that the high end part was a cut down version.
And so it will be in March. Even if nvidia got only 10.000 512SPs at start, it would be dumb to not release this part.
this time we have tesla cards as well as geforce cards. when nvidia talks about gf100 we don't know to which series of cards they refer to. there were rumors that the tesla parts would be higher binned cores than their geforce counterparts.Quote:
Originally Posted by mapel110
the only thing we can do is just wait for the final official specs to be announced.
Wow just wow. Stop taking Nvidia PR as gospel, they're almost as bad as Charlie.Quote:
Originally Posted by ElSel10
Triniboy said it best.
You don't know who's correct :p:
You guys express in every possible way how much you hate nVIDIA and their tricky marketing plot lately, but you still post and fight over who's right or wrong in a meaningless posting war...
Since you're bored and tired of nVIDIA and Fermi, why can't you just stop posting and wait for the launch day ?
GF100 Graphics Architecture White Paper <-- do you expect to read something about Tesla is such a PDF?Quote:
Originally Posted by RaZz!
And Tesla will be slower versions. Because these boards have more vram and need to be able to run 24/7 in computing enviroments.
Different question: do you expect to read something about Geforce in such a PDF?Quote:
Originally Posted by mapel110
kravis @ b3d says he has 470 review numbers:
GTX470 will have a score of 167xx and 73xx, under Performance preset (1280x1048, 1x AA) and Extreme preset (1900x1200, 4xAA) respectively. Which are 13% and 20% faster than GTX285 and is quite competitive when compared to the results of 5870 under Performance preset.
GTX470 will have a wide lead in the benchmarks of the DX10 game, Far Cry 2, with 80fps under the following settings: 1920x1200, 4xAA, 16xAF. However, as there are different benchmarks of Far Cry 2, the data is not ideal for making comparisons with other video cards.
GTX470 can achieve a frame rate of around 30fps during the DX11 Uniengine Benchmark [tessellation] under 1920x1200, with 4xAA and 16xAF, which is apparently better than the 5870.
so are there two benchs existing on the whole freaking world all i hear is heaven and far cry 2 lol
they probably recommend heaven & fc2 in the benchsuite because performance is so good. Kinda like how ati wan'ts people to bench battleforge on the 5xxx.
they removed ''...you may even be able to buy one before anyone else'' from: www.nvidia.com/object/paxeast.html
they must have realized it souded like a joke
Yeah, makes sense.Quote:
Originally Posted by jaredpace
Actually, all the numbers you've posted seem about right. We need some pricing details, though.
Haha, or it just means that you may simply not be able to buy a card then! :DQuote:
Originally Posted by -Sweeper_
Its uber-good, makes Jensen realize he'll have to give it for free.:shocked: Yep, you read that right, for free!:DQuote:
Originally Posted by zalbard
According to these data:
http://img651.imageshack.us/img651/7...70vantagep.jpg http://img685.imageshack.us/img685/6...70vantagex.jpg
Test: Cards 2009 ; A recent overview of the market @computerbase.de
.
.
.
.
HD 5870 is 28% faster than GTX 285 in Vantage Performance, and 37% faster likewise in Vantage Extreme, so if those leaked numbers are correct and coming from a card with the right clock & driver, it really doesn't look good at all for GF 100 prospect. Not to mention that review by computerbase.de is an old one, newer Catalyst will make it even worse. :shrug:
If I say, if ( and that's a BIG IF ) the GTX 470 turns out to be just a bit faster than the 5870, and the GTX 480 about 30-40% faster than the 5870, would you call that a failure ? a success ? or just ok ?
Well, Far Cry2 always performed better on nV hardware ... it looks like nV will only release cherry picked benchmarks.
@benchzowner it would of been a huge success 6 months ago..now id say just ok ..if its anything less..failure.
do you think B is still GF100 or GF104?Quote:
Originally Posted by LordEC911
that was a gift from one of the admins a few years back, heh :DQuote:
Originally Posted by BrowncoatGR
no, he said there will be 512core parts but probably only very very few...Quote:
Originally Posted by ElSel10
yes, so true and so sad... :(Quote:
Originally Posted by annihilat0r
bs! you dont know what your talking about man!Quote:
Originally Posted by mapel110
do some research!
wow... really?Quote:
Originally Posted by -Sweeper_
depends on the price :)Quote:
Originally Posted by BenchZowner
if gtx480 costs 400$ and is 35% faster than a 5870 thats VERY nice!!!
if it costs 35% more for being 35% faster, then its ok... (~533$)
if the 470 costs 250$ and is between a 5850 and 5870, thats VERY nice!!!
if it costs about the same as a 5850 (~300$) and is as fast as a 5870, then thats ok...
That's not bad at all. At those performance margins prices settle nicely. But then again, the nvidia vomit will come back with a vengeance.Quote:
Originally Posted by BenchZowner
For us the consumer it depends on the cost, if the GTX 480 is around 30% faster then a 5870 and costs 40% more its a mild success.Quote:
Originally Posted by BenchZowner
For Nvidia Fermi is simply to late to be a success for them, its become a Joke, my friends who know nothing about Video Cards, now know about Fermi and how late and how much of a joke it has become.
Fermi has been and will continue to be a PR problem for Nvidia unless it rocks the world, even a 30% speed increase over 5870 is not going to make people forget that its 2010 and we wanted Fermi in 2009.
I was planing on going to PAX anyway, and I am rather excited to see what Nvidia has to offer, not because I expect Fermi to be amazing, but I want to see them talk there way out of preventing me from buying one when I am there.
I don't think the difference between the 470 and 480 will be that big, but anyway, if it would indeed be 30% - 40% faster than the 5870 then yes it could be considered a success, simply put, the ati refresh would not be able to counter it then, so it would be a success for nvidia, but of course we would loose as it would mean high prices.
We need the 470 to be priced around 400, not more. It has got to be able to beat 5870 in price/performance, else failure. If 480 = ~30%>5870 And costs 30% more than 5870, then all this wait would have been wasted b/c 5870 price will not go down fast enough.
30% in most games over a 5870 is pretty damn impressive and I'm more leaning towards AMD side. 20-30% meh and <20% = Fail.
Not to mention AMD is more than capable of slashing prices after all of this dominance.Quote:
Originally Posted by kadozer
Imagine AMD drops prices by $100 the day that 470 launches.
That wouldn't just be rubbing salt in the wounds. That would be:
- Rubbing salt in the wounds
- Add some vinegar
- A bit of alcohol
- A flamethrower to blend everything together
- A pinch of salt on top
Can AMD do it?
they can and still make a nice margin on their cards afaik :DQuote:
Originally Posted by damha
but whether they will do it is another question...
if you consider that amd might still be limited by tsmcs 40nm wafer output, and nvidia is even more volume restricted... it doesnt make sense to cut prices as it barely affects marketshare or sales volume : /
I downgraded from a GTX 260 to a Evga GTS 250 1gb and the difference is only marginal, sometimes 6fps, for me there won't be a need to upgrade graphics card until new consoles are out since they're the one holding back progress.
I'd consider 30 percent faster a success, if it is priced right.
Imho the question is: does AMD want that? I think they are happy the way prices are and NVIDIA (obviously) doesn't have much room to go down with the price.Quote:
Originally Posted by damha
I think prices will stay quite high compared to last generation.
Honestly, I think if 480 is 30-40% faster than 5870, that means it will literally touch 5970 performance, often match it.Quote:
Originally Posted by BenchZowner
not gonna happen.
20% faster is a success at a competitive price.
PS: If 480 was 35% faster than 5870, nvidia would not stop talking about it...
Yes we do. It was disclosed in the whitepaper that GF100 denotes GeForce and Tesla is something else.Quote:
Originally Posted by RaZz!
Couldn't have said it better.Quote:
Originally Posted by trinibwoy
Yes, I am sure the master plan is to produce a chip that is never actually used. How did I not see that before?Quote:
GF100 is a chip. GTX 480 is a part based on that chip that may or may not have all SM's enabled.
Yes, that is for Tesla. It's been confirmed by Nvidia from several sources that the high end GeForce part has the full 512 SPs. Period. There isn't any argument here and you're not doing anything for your reputation by trying to create one.Quote:
You know how the top announced Tesla SKU only has 14 of the SM's enabled though it's based on the same chip? Why is such a simple thing so difficult for you to understand?
Yeah, they removed the vague wording and have instead revealed the GTX470 and 480 launch date outright.Quote:
Originally Posted by -Sweeper_
i think that's being very optimistic, if the gtx470 can beat the 5870 then i can see them asking nearly $500 for it. that makes me cringe at the price of a gtx480... we may find ourselves back to the price gouging days... i'm certain nvidia will price the cards according to their final performance numbers, so we might not see much of a price war. :(Quote:
Originally Posted by saaya
that's because you're running running the latest games with AA at 1920x1200 or higher resolution ;)Quote:
Originally Posted by taskforce