View Full Version : R520 Benchmark on 3dmark05!
sandeep
06-07-05, 11:57 PM
I hearing about R520 in trouble, I think I am gonna get 2*6800GT and be happy. Although I was really hoping to get a 32 pipeline card in SLI.
I would wait and see first before buying anything. The G70 launch is just 2 weeks away.
AthlonXP1800
06-08-05, 02:00 AM
I would wait and see first before buying anything. The G70 launch is just 2 weeks away.
Quite.... now it just 13 days away. :angel:
PeterJensen
06-08-05, 02:44 AM
AthlonXP1800 : ohhh time flyes :) Btw i live here in Denmark and well the hardware is always more expensive over here. BUT i got 4000dkr (800$) just waiting for at G70 :D
Zelda_fan
06-10-05, 09:08 AM
LOL cmon man...Nvidia ain't really got much going on this coming gen when you really think about it. I'll probably get a G70 when it comes...but
ATI has more features coming SM3.0...crossfire...There's nothing new about G70 except its a faster 6800...pretty much.
ATI is not going to release a 32 pipe card to start with...especially if Nvidia doesn't... I hate to say this, but I dont think Nvidia is suprising anyone this generation according to the specs released for it...and the fact that its supposed to be here in ~1 month...suggests to me that its nothing but more pipes. Nvidia has no idea what kind of numbers ATI is getting now with SM3.0, among other things.
This is IMO, as fan of both sides...I see it going something like this:
- Nvidia releases G70...best card out at the moment...but many people with SLI rigs dont see much point in upgrading cause the card doesn't offer enough features over the 6800's. The Nvidia people without SLI buy it up..
- ATI releases Crossfire to take steam out of G70 launch...people wanting a bit better performance buy another x8x0 card and are happy.
- 1 month later...ATI releases R520..which beats the G70 by a bit...
- Nvidia releases new drivers increasing performance somehow...then scrambles together either a 32 pipe version or some dual GPU version that costs $2000 with a gig of ram...that no one will buy...to combat the r520
- ATI follows suit and releases a 32 pipe version and ends up a bit faster...
but anyways...this gen is trivial...NEXT gen after this one, with unified shaders is where it will start to get interesting.
precisely why I think NV is going to end up having faster cards out. ATI is adding a TRUCK LOAD of new features into the R520 - features that nVidia dosen't need to develope because they alrady have. nVidia can concentrate on speed while ATI has to play catchup.
Increasing pipes may not be innovative, but if it makes the card a lot faster who cares?
I really doubt ATI is going to sell much crossfire. When it comes down to it, SLI is just as fast (much faster when G70 comes out) but costs less. Why would anyone buy such a thing?
nVidia would be completely retarted to set a MSRP of 2000 for a card. I would think at the most they would charge 500 for the 32 pipe card.
ChrisRay
06-10-05, 06:42 PM
Sure..ATI is just coming out with SM3.0 now...but can you say they missed out on much? How much more do you have to criticize them for not having it right now when 3 games actually use features of it? How do you know what ATI's SM3.0 performance will be like? Surely ATI will have used the extra time to make better implementations for SM3.0 and to add a couple additional features.(same thing is true for nvidia, I know) Some people act like ATI never knew of SM3.0 before Nvidia shocked the world with it.
Looking at each companys tech though, should I say anything about Nvidia playing catch up with unified shaders? Why are they not in the RSX? Surely Nvidia will adopt when WGF2.0 comes around right? Well ATI's doing the same thing now...
First of all. Unified shaders arent even required for WGF 2.0. Secondly Nvidia specifically said they have no intention of implementing due to their "vision" of the GPU pipeline. Unified shaders main benefit is to make sure there are no idling units in the graphic pipeline. They are not an improvement in featureset.
I wasnt criticizing ATI for not having SM 3.0. There are plenty of other places and discussions for that. But criticising Nvidia for being first and being ahead as far as DirectX goes? And only needing a refresh? What more do you want from the company? There's not much they can do until its closer to WGF 2.0 release.
If you would read what I'm saying...I'm saying that Nvidia is NOT going to have any special features that make it unique this time...meaning its squarly performance vs performance...not.."Oh I'll get this one cause it has SM3.0 and the other one doesn't." I'm not saying anything about them not having anything new...just that people shouldn't expect any drastic changes that make them stand out over ATI..Vice versa for ATI, but ATI people finally will have SM3.0 and Crossfire. Nvidia is not(as I know of) bringing out anything that is saying "buy us instead of them, cause we have this" this gen.
Thing is. I am reading what you're saying. And no offense, I find it short sighted criticism. And wrongfully damning.
Vagrant Zero
06-11-05, 12:51 AM
There's an old saying, don't fix it unless it's borked. Current architectures have worked so far, if someone wants to throw a rock in the puddle and claim Unified is better I want some damned proof first. And not PR speak either, I want benches. Until then, oh look marketing, don't step in the marketing.
Brick_Top
06-11-05, 03:09 PM
There's an old saying, don't fix it unless it's borked. Current architectures have worked so far, if someone wants to throw a rock in the puddle and claim Unified is better I want some damned proof first. And not PR speak either, I want benches. Until then, oh look marketing, don't step in the marketing.
So you mean ATI is using the xbox360 as a guinea pig for an unified shader architecture?
XMAN52373
06-11-05, 10:05 PM
"For what we know Nvidia can not mach this with G70." As much as this is Inquirer stuff, but in my capacity, I have found them to be correct!!
I am being very presumptuous.
Well I wouldn't liked to be milked by nVidia for a G70 and later for a G80 within a mere span of 6months.
If ATI offers a full blown setup now, it should be fun cos nVidias G70 will be inferior. Not only that, but the fact CROSSFIRE with 2*X850XT PE beat 6800ULTRA SLI!!!!!! in D O O M 3 (0-o) :clap2: What will ATI do in other games ;)
Head over here to surprise yourself--->http://www.tweaktown.com/document.php?dType=article&dId=773
Nvidia should facilitate to get their G70 out ASAP for it to earn a dime orelse, R520 could potentially still the show.
I hope I am keeping my feet within the boundaries of 'Discussion', negating the fact to be stereotyped. ;)
Its a shame the numbers they used for 2 6800Us in SLI are toal and utter BULL****! The proof is here at Toms.
http://graphics.tomshardware.com/graphic/20041222/vga_charts-06.html
I bet they got paid by ATI to use bogus numbers for 2 6800Us in SLI to help promote crossfire.
Even more proof of the BS of that link!
http://www.hardocp.com/article.html?art=Njk2LDM=
XMAN52373
06-11-05, 10:09 PM
If you would read what I'm saying...I'm saying that Nvidia is NOT going to have any special features that make it unique this time...meaning its squarly performance vs performance...not.."Oh I'll get this one cause it has SM3.0 and the other one doesn't." I'm not saying anything about them not having anything new...just that people shouldn't expect any drastic changes that make them stand out over ATI..Vice versa for ATI, but ATI people finally will have SM3.0 and Crossfire. Nvidia is not(as I know of) bringing out anything that is saying "buy us instead of them, cause we have this" this gen.
Sure..ATI is just coming out with SM3.0 now...but can you say they missed out on much? How much more do you have to criticize them for not having it right now when 3 games actually use features of it? How do you know what ATI's SM3.0 performance will be like? Surely ATI will have used the extra time to make better implementations for SM3.0 and to add a couple additional features.(same thing is true for nvidia, I know) Some people act like ATI never knew of SM3.0 before Nvidia shocked the world with it.
Looking at each companys tech though, should I say anything about Nvidia playing catch up with unified shaders? Why are they not in the RSX? Surely Nvidia will adopt when WGF2.0 comes around right? Well ATI's doing the same thing now...
Regardless of what happens this gen....like I've said before..Microsoft and ATI are in bed together right now...and with WGF2.0, unified shader's and XNA, Its looking like ATI will be in a good position with microsoft when longhorn comes..
guys...ATI has the tech...it always has...always will...their only problem is getting these cards out in good fashion. If they can supply this card in great quantities, then they should have no problems.
As the saying goes..and I agree with it..
"Nvidia's driver/software/PR teams have ALWAYS saved their hardware's butt."
ATI has not always had the tech. Rage series and first 2 runs of the radeon 7xxx and 8xxx lines blew chunks compared to Nvidia.
Subtestube
06-11-05, 11:11 PM
Besides which, the NV40 was certainly technically more advanced than the R420. My understanding is that the tech will be equivalent between ATi and NV this time around - let's wait for the benchies and the image quality comparisons, and we'll see how the two fare. It'll be a genuinely interesting race, as we know very little about how either new card performs
a) per clock and
b) overall.
If there is some featureset difference, we'll all know soon.
First of all. Unified shaders arent even required for WGF 2.0. Secondly Nvidia specifically said they have no intention of implementing due to their "vision" of the GPU pipeline. Unified shaders main benefit is to make sure there are no idling units in the graphic pipeline. They are not an improvement in featureset.
ugh...ok.
I never said Unified Shaders are a requirement for any featureset...I said that unified shaders will be optimized through WGF2.0 when it gets here, and that IMO ATI will be in a better position when that happens! I know what the point of unified shaders is...perhaps you can explain to me what the point of sticking with seperate Pixel pipelines/vertex units is? What has Nvidia so enlightened, that they would stick to that approach? I'm not being sarcastic...I wanna know the upside to this. Specifically when we know that much of the hardware isn't being utilized to the fullest.
Why is it so wrong for me to think that unified shaders is the way of the future? And why cant I criticise nvidia's approach to this? Am I doing something wrong? Do you not like to question their approach? When do we start optimizing the hardware we have so that its used to its fullest...
I wasnt criticizing ATI for not having SM 3.0. There are plenty of other places and discussions for that. But criticising Nvidia for being first and being ahead as far as DirectX goes? And only needing a refresh? What more do you want from the company? There's not much they can do until its closer to WGF 2.0 release.
I agree, this is the wrong thread for this, sorry...but, where was I criticising Nvidia over being first with SM3.0? I simply said that they hyped it up like it was needed at the time... I wasen't "damning" Nvidia for not bringing anything new and praising ATI for bringing SM3.0...rather saying that it appears as if ATI is improving more over their last line up than Nvidia is...
I'm not expecting much from Nvidia this time around, I just said they ain't really got much new coming over the 6800 when you really think about it. Any hardware entheusiest knows there's always more that could be done.. The G70 is a new core. For Nvidia, what would be the point in bringing out a new core this time around if all your really doing to increase speed is adding more PP/VU's? The point is.. I know Nvidia is going to be adding improved features to their cards, the new core makes it obvious.
let me quote myself..
If you would read what I'm saying...I'm saying that Nvidia is NOT going to have any special features that make it unique this time...meaning its squarly performance vs performance...not.."Oh I'll get this one cause it has SM3.0 and the other one doesn't." I'm not saying anything about them not having anything new...just that people shouldn't expect any drastic changes that make them stand out over ATI..Vice versa for ATI, but ATI people finally will have SM3.0 and Crossfire. Nvidia is not(as I know of) bringing out anything that is saying "buy us instead of them, cause we have this" this gen.
Thing is. I am reading what you're saying. And no offense, I find it short sighted criticism. And wrongfully damning.
Think what you will. I know you're a smart person...but you're really not even giving me a chance here. You're jumping on everything I say..thinking I'm bashing nvidia all the time or something, when what I'm really doing is comparing the companies by how much they improve themselves over the previous gen.
Is it all because of my first post here? :p
There's an old saying, don't fix it unless it's borked. Current architectures have worked so far, if someone wants to throw a rock in the puddle and claim Unified is better I want some damned proof first. And not PR speak either, I want benches. Until then, oh look marketing, don't step in the marketing.
Dont fix it unless its borked?? ok then...you have fun paying $500-600 for a card thats working at about 60% effeciency half the time... With people like you...the way hardware works would never change..just faster versions of the same old same old inefficent crap. The benefits of Unifiedshaders are obvious..in situations where either the vertex OR pixel shaders is bottlenecked. In games that dont stress either...you wont see a difference. Thats not marketing...thats just common sense.
ATI has not always had the tech. Rage series and first 2 runs of the radeon 7xxx and 8xxx lines blew chunks compared to Nvidia.
Damn... Yea, I worded that wrong for sure, lol. I agree 100%.
Besides which, the NV40 was certainly technically more advanced than the R420. My understanding is that the tech will be equivalent between ATi and NV this time around - let's wait for the benchies and the image quality comparisons, and we'll see how the two fare. It'll be a genuinely interesting race, as we know very little about how either new card performs
a) per clock and
b) overall.
If there is some featureset difference, we'll all know soon.
Thank you...this post is the truth. Yes, technologically speaking the NV40 is more advanced than the R420 because it supported more features. Now with that field leveled...it will be interesting to see which architecture is more effecient.
ChrisRay
06-12-05, 02:00 AM
ugh...ok.
I never said Unified Shaders are a requirement for any featureset...I said that unified shaders will be optimized through WGF2.0 when it gets here, and that IMO ATI will be in a better position when that happens! I know what the point of unified shaders is...perhaps you can explain to me what the point of sticking with seperate Pixel pipelines/vertex units is? What has Nvidia so enlightened, that they would stick to that approach? I'm not being sarcastic...I wanna know the upside to this. Specifically when we know that much of the hardware isn't being utilized to the fullest.
Why is it so wrong for me to think that unified shaders is the way of the future? And why cant I criticise nvidia's approach to this? Am I doing something wrong? Do you not like to question their approach? When do we start optimizing the hardware we have so that its used to its fullest...
Err the point of sticking too seperate vertex units is because theoretically they can achieve far greater potential than a unified shader setup. But again that is entirely dependent on the pipeline setup. But that is dependent on the units as well. The big reason to go unified shaders is to save transistor count and have units that dont go to waiste. Geometry/Pixel engines share their role. So if there is not a lot of geometry processing. The units dont go to waiste. There is absolutely no reason that a seperate vertex/pixel engine setup would have to be slower. Perhaps just less efficient. If Nvidia can achieve greater performance by not sharing pixel and geometry units and still maintain a comparable featureset. Then there absolutely nothing wrong with that.
Unified shaders are not going to be a considered a "Feature" set from a WGF 2.0 standpoint. Nor is running them going to give you some kind of gaurenteed advantage in DirectX.
I agree, this is the wrong thread for this, sorry...but, where was I criticising Nvidia over being first with SM3.0? I simply said that they hyped it up like it was needed at the time... I wasen't "damning" Nvidia for not bringing anything new and praising ATI for bringing SM3.0...rather saying that it appears as if ATI is improving more over their last line up than Nvidia is...
I'm not expecting much from Nvidia this time around, I just said they ain't really got much new coming over the 6800 when you really think about it. Any hardware entheusiest knows there's always more that could be done.. The G70 is a new core. For Nvidia, what would be the point in bringing out a new core this time around if all your really doing to increase speed is adding more PP/VU's? The point is.. I know Nvidia is going to be adding improved features to their cards, the new core makes it obvious.
How do you know whats in the pipeline at this point? How do you know what Nvidia has planned to add? Lets look at it from a historical standpoint.
TNT -- Geforce
First geometry assisting engine
Geforce ---> Geforce 2
Additional TMU per pipe. Trilinear Filtering finally for free.
Geforce 2 --- Geforce 3
Multi Sampling, Programmable shaders, Improved Ramdac, Nview.
Geforce 3 -- Geforce 4
PS 1.1 to PS 1.3 support. Additional Vertex units, Quincunx filter added to ramdac. Memory Controller improvementsand impropved Z-Compression. Dual Ramdac.
Geforce 4 --> Geforce FX ((Nv3x))
Multisampling Bandwith Compression, PS 1.3 to PS 1.4/2.0, Improved vertex performance, ((Cant put a number of FX vertex engine) ((perhaps not the best chip ever released but still significant improvements)) Double Sided Stencil Acceleration.
NV3X --- > Nv35
256 Memory Bus, replaced integer units to a FP ALU. Introduction, Improved Shadow Volume support.
Geforce FX --- Geforce 6 series
RGMS @ 4x, PS 2.0 ---> 3.0, OpenEXR FP16 Blending support, SLI, Angle Dependent AF ((if you can call that an improvement)) Dual Shader ALU's per pipe.
Geforce 6 ----> Geforce 7.
??. But judging from past history. We can still assume there are going to be architecture improvements. Many guesses can go from improvements to LoD calculations in AF, Improved pipeline/ALU efficiency with less texture dependency. OpenEXR blending with AA. Right now all this is up in the air. Just because their are "2" SM 3.0 cards out doesnt
The point to be made here is that while there may not be huge strides in the DX featureset. We dont "know" what the actual G70 Pipeline holds. Since you're jumping to conclusions and clamoring on about how SM 3.0 wasnt needed yet and when ATI gets here it'll be the next big thing. ((Which is ridiculous since the NV4x feature set is being adopted quicker than the DX 9.0 baseline feature set was)). I really dont see how any could argue that Nvidia made a mistake marketing and using its superior featureset when we can already see that it's already implemented several games and many new applications to come. If SM 3.0 wasnt a good enough feature on the Nv4x the FP16 blending was. And they have seen some realtime use "now". And they're going to see alot more use in the near coming future. Heck even ATI"s new lost coast level is going to take advantage of FP blending. Actually you should probably be thanking Nvidia for its SM 3.0 marketing. Because by the time ATI"s chipset comes out with SM 3.0 compliance. They can bank on software already being compatible due to all that "nvidia" marketing. SM 3.0 is a standard. No one ever argued it wasnt. The G70 as a refresh is already far more interesting than any second refresh product I have seen in the past. Improved pixel pipeline layout and improved vertex engine layout. And judging from what people have been seeing in the 77.13 drivers. There's likely to be more than just that.
I'm onto your case because I find your opinions to be one sided. And I think you have fallen way too far into the hype of "Unified Shaders". SM 3.0 will be here for a least another year. And its the big thing that both major players are going to adopt. You should consider it important and you should pay very close attention to how both vendors run FP32, Dynamic and Static branching. This year is likely to be very interesting once we see how ATI implements its DX 9.0C compliant chipset.
Please. Do me a favor. I heard you the first time. I dont need to see you quote yourself over and over. I still think it's wrongfully damning and overly speculative. Then again. I will say this again. Trying to name a winner this early. When both hardwares are still officially vaporware. Seems like an act of grasping at straws. This is why I so hate rumor forums. Its nothing but fanboys speculating an arguing. Promoting their side of the agenda. Why dont you just wait until the hardware is out before drawing silly conclusions? This is my final post on this subject. I got better things to do than argue about vaporware.
P.S And forgive my annoyance. My biggest pet peeve is people who think they know everything or anything about next generation products and say it as gospel.
Unified shaders are not going to be a considered a "Feature" set from a WGF 2.0 standpoint. Nor is running them going to give you some kind of gaurenteed advantage in DirectX.
I never said they were for crying out loud! I do believe however that eventually there will be games programmed that will take advantage of unified shaders....
How do you know whats in the pipeline at this point? How do you know what Nvidia has planned to add?
I dont know what Nvidia has planned for adding features...but you're the one that said there's not much they could do until a new featureset arrives....
The point to be made here is that while there may not be huge strides in the DX featureset. We dont "know" what the actual G70 Pipeline holds. Since you're jumping to conclusions and clamoring on about how SM 3.0 wasnt needed yet and when ATI gets here it'll be the next big thing. ((Which is ridiculous since the NV4x feature set is being adopted quicker than the DX 9.0 baseline feature set was)). I really dont see how any could argue that Nvidia made a mistake marketing and using its superior featureset when we can already see that it's already implemented several games and many new applications to come. If SM 3.0 wasnt a good enough feature on the Nv4x the FP16 blending was. And they have seen some realtime use "now". And they're going to see alot more use in the near coming future.
I never said there was any problem with nvidia having SM3.0. What you're saying is that I should be criticising ATI for not having it when I simply cant because I didn't feel the software needed it at that time...
I can't help the fact that ATI is adopting it now..at a time where its starting to be used in games. How long was the 6800 out before anything took advantage of SM3.0 excluesively(without patch) besides the nvidia demos used to show it off?
Heck even ATI"s new lost coast level is going to take advantage of FP blending.
Its not ATI's level...and ofcourse FP blending is a nice feature... Last I heard the R520 will have it.
Actually you should probably be thanking Nvidia for its SM 3.0 marketing. Because by the time ATI"s chipset comes out with SM 3.0 compliance. They can bank on software already being compatible due to all that "nvidia" marketing. SM 3.0 is a standard. No one ever argued it wasnt. The G70 as a refresh is already far more interesting than any second refresh product I have seen in the past. Improved pixel pipeline layout and improved vertex engine layout. And judging from what people have been seeing in the 77.13 drivers. There's likely to be more than just that.
Yes, I do thank Nvidia's marketing team for getting ubisoft to code a SM3.0 setting and no 2.0 setting in Splinter cell. Makes the adoption process that much quicker.
Thats the way its always worked my friend....Nvidia comes out with some exciting new technology...pimps it out...markets the hell out of it...than other companies come and reap the spoils right when the time is right. Seems like a smart idea to me.
I'm onto your case because I find your opinions to be one sided. And I think you have fallen way too far into the hype of "Unified Shaders". SM 3.0 will be here for a least another year. And its the big thing that both major players are going to adopt. You should consider it important and you should pay very close attention to how both vendors run FP32, Dynamic and Static branching. This year is likely to be very interesting once we see how ATI implements its DX 9.0C compliant chipset.
Perhaps I've bought into the hype about unified shaders..but I believe that its going to be adopted fairly quickly, and Nvidia will be right there when the time is right...then when they come out with a unified shader card...what are you gonna say? "Oh...unified shaders are where its at now." Maybe...maybe not. Maybe Nvidia will never have unified shaders...I dunno, as long as they have better performance..I dont care..I'll buy their cards.
Please. Do me a favor. I heard you the first time. I dont need to see you quote yourself over and over. I still think it's wrongfully damning and overly speculative. Then again. I will say this again. Trying to name a winner this early. When both hardwares are still officially vaporware. Seems like an act of grasping at straws. This is why I so hate rumor forums. Its nothing but fanboys speculating an arguing. Promoting their side of the agenda. Why dont you just wait until the hardware is out before drawing silly conclusions? This is my final post on this subject. I got better things to do than argue about vaporware.
Perhaps I am over speculating... I agree trying to name a winner right now is stupid.
I've told you again and again..I'll buy whichever card is better. Thats why I'm so excited about this gen because they will both support the same feature set. I can get a better grasp of which company has the better implementation.
P.S And forgive my annoyance. My biggest pet peeve is people who think they know everything or anything about next generation products and say it as gospel.
:angel:
I'm saying that Nvidia is NOT going to have any special features that make it unique this time...
Well they won't have SM3.0 to dangle in front of Ati, but what evidence do you have that they wont have any special features to differentiate themselves?
Links are fine :p
Well they won't have SM3.0 to dangle in front of Ati, but what evidence do you have that they wont have any special features to differentiate themselves?
Links are fine :p
I know man. I hope they have something kickass that no one knows about..especially ATI. It goes both ways though. It would be bad for Nvidia to assume that ATI is just playing catchup with them and not really expand on their hardware's features...thats all I was ever saying. It looks to me like they are just tweaking and what they already have...which is fine!
I just dont see how people are saying that ATI is playing catchup with nvidia, when ATI made a consious decision to not support SM3.0 with their X800 series. It doesn't mean that ATI is going to start blind...I'm sure they haven't been wasting their time not experimenting with it. No one has any idea what ATI has in store either. Yet they say Nvidia will probably be better cause they came out with it first.
What I'm basing my theories off of is the fact that Nvidia already has a SM3.0 architecture out there, which means they're more likely to build off what they have, whereas ATI is coming with something new, and is sure to be making it better than what their competitors currently have.
And in regards to that quote...I meant they aren't really going to have any feature set advantages over ATI...their hardware support for features may be a different story.
I just dont see how people are saying that ATI is playing catchup with nvidia, when ATI made a consious decision to not support SM3.0 with their X800 series. It doesn't mean that ATI is going to start blind...I'm sure they haven't been wasting their time not experimenting with it. No one has any idea what ATI has in store either. Yet they say Nvidia will probably be better cause they came out with it first.
I think people say that b/c they didn't have a SM 3.0 product the last time around. So, this time around it'll be ATI's first attempt at SM 3.0 but it'll be nVidia's 2nd. Products are usually refined over time, normally getting better each time (whether it's cars, cpus, video cards, appliances, drivers, software apps, etc.).
I think people say that b/c they didn't have a SM 3.0 product the last time around. So, this time around it'll be ATI's first attempt at SM 3.0 but it'll be nVidia's 2nd. Products are usually refined over time, normally getting better each time (whether it's cars, cpus, video cards, appliances, drivers, software apps, etc.).
Yea.
This is a purely hypothetical question as we have no hardware to test...but do you think Nvidia's 2nd attempt will be better than what ATI has with the R520? And also...with the architectures being so different...its hard to compare spec for spec...but as of now, I'm leaning towards the R520 being the faster card.
...but as of now, I'm leaning towards the R520 being the faster card.If ATI releases after nvidia like everyone expects, they'd better hope they're faster (or very, very close) or else they could be in trouble. I think in general, people expect products released later to be better and faster. If it doesn't happen, it's kind of a feeling like...I waited for this??? That's why I think the PS3 had better end up being more powerful than the xbox 360. If you release that much later, they had better be x number of months better or else people will be disappointed. Not really wanting to start another console discussion but I felt it was related since they won't be launching at the same time either.
If ATI releases after nvidia like everyone expects, they'd better hope they're faster (or very, very close) or else they could be in trouble. I think in general, people expect products released later to be better and faster. If it doesn't happen, it's kind of a feeling like...I waited for this??? That's why I think the PS3 had better end up being more powerful than the xbox 360. If you release that much later, you had better be x number of months better or else people will be disappointed. Not really wanting to start another console discussion but I felt it was related since they won't be launching at the same time either.
Well not really, remember Gf FX launched much later to r300.. and got..well..you knw wat its called..hmm...|2@P3|)
I'm not sure I follow you since I think that just proves what I was saying. If the product is released later, people expect it to be better, faster, etc. If it's not, it spells trouble.
I'm not sure I follow you since I think that just proves what I was saying. If the product is released later, people expect it to be better, faster, etc. If it's not, it spells trouble.
lol yea.
I should also state that I'm aware the same thing could happen to ATI that happened to Nvidia with the NV3.0. I hope it doesn't..but its a possibility. If it does...then I'm going to go with the g70.
I dont know why, but i hav this feeling, ATi are not completely satisfied with the r520. so they havent even hinted, abt the r520 and its performance. all this "INFO" we hav on r52x is just made up by people. also they hav the pressure of xfire. they hav said a lot about xfire, but werent able to prove any of those.
Also i remember, the r5xx core was developed, fabricated, and tested much earlier to development on g70..it was in november 2k4. and they were even satisfied with it. but as usual, its delayed..
So i am favoring nv for delivering a better product this time..
But we've seen in the past, ATi are extremely competitive, when it comes to grabbing top performance crown.
it all depends actually. the product available in $400 price, with best performance/price ratio will be the winner for me.
It was 6800GT earlier, and 9700pro before gf6, ti4200 (was $250) but still..
"For what we know Nvidia can not mach this with G70." As much as this is Inquirer stuff, but in my capacity, I have found them to be correct!!
I am being very presumptuous.
Well I wouldn't liked to be milked by nVidia for a G70 and later for a G80 within a mere span of 6months.
If ATI offers a full blown setup now, it should be fun cos nVidias G70 will be inferior. Not only that, but the fact CROSSFIRE with 2*X850XT PE beat 6800ULTRA SLI!!!!!! in D O O M 3 (0-o) :clap2: What will ATI do in other games ;)
Head over here to surprise yourself--->http://www.tweaktown.com/document.php?dType=article&dId=773
Nvidia should facilitate to get their G70 out ASAP for it to earn a dime orelse, R520 could potentially still the show.
I hope I am keeping my feet within the boundaries of 'Discussion', negating the fact to be stereotyped. ;)
Huh... I'm sorry to say this, but Hardinfo and Tweaktown has completely borked Doom 3 tests in the Crossfire preview. I get like 15fps more with single 6800u and same timedemo + config file than Hardinfo and Tweaktown. They had also faster system than I have right now (2.8@3.2 + 1GB ram + 77.xx drivers). Imo, 6800U SLI should get more frames ....
I get even more frames (+3fps) with 3mood (everything is maxed, ultra quality specular and offset mapping) than Hardinfo without 3mood :)
Download Hardinfo demo + config file (http://www.hardinfo.dk/art/site/pic/download/timedemo/hardinfo_doom3_benchmarking.zip)
vBulletin® v3.7.1, Copyright ©2000-2012, Jelsoft Enterprises Ltd.