New rumor about ATI Southern Islands

Printable View

Show 100 post(s) from this thread on one page

08-29-2010, 11:54 PM
flopper

Quote:

Originally Posted by Carfax

If AMD genuinely desires to address the HPC market, they will have to invest more transistors into their designs for dedicated HPC function....which will mirror what Nvidia has done with Fermi.

Of course, the AMD fan boys will herald this as the greatest thing ever even if the power envelope is increased dramatically, much as what happened with Fermi :rolleyes:

Fusion as a concept and the aim for power/price/performance strategy will allow them to come up with stuff that suits that concept of a HPC tech card down the line.
I cant think they be skipping a growing new market as we move towards a union between different structural elements, combining for example a videocamera with a normal camera is now on the way, you buy both in the same toy as the technology now have moved far with that.
Like the phone that is now, music, video, chat, and well a phone.

social networking is increasing, at a rate that is astounding, and software and hardware needs to met up there as that continues.
If I be hired by amd I advice them to get going there with their commercials even more.

So, our hardware, video and cpu, and other things is getting soon to a point where we can have a combined element, and fermi is just powerhungry due to bad design, first generation tends to be that.
I expect that Nvidia will improve the power issue a lot in the upcoming revisions they do.
08-29-2010, 11:54 PM
STEvil

Hey carfax. Multi-quote.
08-29-2010, 11:55 PM
Florinmocanu

It is a failure, compared to the 8000s series. It's a "meh" architecture when gaming is concerned because it takes a huge die and a lot more TDP to surpass the smaller and cooler Cypress chip. that is, perf/mm.
08-30-2010, 12:16 AM
madcho

Quote:

Originally Posted by Florinmocanu

It is a failure, compared to the 8000s series. It's a "meh" architecture when gaming is concerned because it takes a huge die and a lot more TDP to surpass the smaller and cooler Cypress chip. that is, perf/mm.

Cypress is far from full optimised. Cayman will be better to that. So i think too, fermi is a total failure, like was old 512bits HD2900
08-30-2010, 12:35 AM
spursindonesia

Darn, as i've said before, incoming FUDs & fanbois talks are soooo predictable, and now the thread is going fubar as it is. :rolleyes: Just say i don't fu<kin believe the numbers and moving on, ATi is da suxxorz that incapable of such feats, that's still very much respectable in my view. :cool:
08-30-2010, 03:54 AM
rozzyroz

Quote:

Originally Posted by Dan_c

Makes my english look fantastic!:rofl:

Finger's up tho, keep'em comin':up:

classic :D

i think you mean thumbs up... fingers up means something completely different :yepp:
08-30-2010, 04:37 AM
LightSpeed

Quote:

Originally Posted by Carfax

:ROTF:

Any long term Nvidia user will tell you that it's nothing for Nvidia to squeeze about 15% extra performance (average) or more out of their GPUs between the introductory drivers, and the fully optimized drivers; which may take about a year or more to realize..

But the current drivers are not introductory drivers. Fermi already received a performance boost with the latest drivers. Also, why are you ignoring the fact the cayman is a new architecture as well, that can be tweaked for more performance with newer drivers too :shrug:
08-30-2010, 04:44 AM
spursindonesia

Quote:

Originally Posted by LightSpeed

But the current drivers are not introductory drivers. Fermi already received a performance boost with the latest drivers. Also, why are you ignoring the fact the cayman is a new architecture as well, that can be tweaked for more performance with newer drivers too :shrug:

For fanbois, the road only goes one way. ;) ...... :ROTF:
08-30-2010, 04:59 AM
Carfax

Quote:

Originally Posted by flopper

So, our hardware, video and cpu, and other things is getting soon to a point where we can have a combined element, and fermi is just powerhungry due to bad design, first generation tends to be that.

Unless you're a semiconductor engineer, how could you possibly know that Fermi is power hungry due to "bad design?"

I don't think most people here understand that Fermi is the first TRULY computational GPU. All those things you said about convergent technologies into one neat package, are made manifest in Fermi....in the sense that Fermi has many CPU like qualities such as being fully IEEE compliant, support for C++, having dedicated caches and support for much greater VRAM.

Quote:

Originally Posted by STEvil

Hey carfax. Multi-quote.

:D

Quote:

Originally Posted by Florinmocanu

It is a failure, compared to the 8000s series. It's a "meh" architecture when gaming is concerned because it takes a huge die and a lot more TDP to surpass the smaller and cooler Cypress chip. that is, perf/mm.

The very fact that a GPU like Fermi, which isn't even a pure gamer chip can outperform Cypress (which is dedicated towards gaming) speaks volumes in and of itself.

And as I mentioned before, Fermi's performance is increasing with every driver update. I've noticed a lot of differences between the various driver revisions since I've owned mine.

It wouldn't surprise me if Fermi ends up being 20% faster than it was on introduction, simply due to driver optimizations.
08-30-2010, 05:13 AM
Carfax

Quote:

Originally Posted by LightSpeed

But the current drivers are not introductory drivers. Fermi already received a performance boost with the latest drivers.

Yes, the current drivers are no longer introductory and have Fermi specific optimizations. But it's only been a few months since the release of Fermi, and they've already gotten large performance increases.

It usually takes a year or more to really milk the architecture for all it's worth, so more performance increases are certainly coming.

Quote:

Also, why are you ignoring the fact the cayman is a new architecture as well, that can be tweaked for more performance with newer drivers too :shrug:

I wouldn't say I was ignoring it. It's just that even though Cayman is a new architecture, it will still share very many similarities to it's predecessors.

Since they migrated towards a unified shader architecture, ATI's strategy has been to simply double all the resources on new designs. I think the only thing that ATI didn't double with the R800 was the memory interface, which remained at 256 bit.

And since Cayman won't be getting a die shrink, it seems like it will be a more refined version of the R800, rather than a next generation architecture.

Anyway, driver optimization is an ongoing affair with no end in sight. Cayman will undoubtedly receive performance improvements via driver tweaks, but I doubt it will get as much as Fermi, since Fermi was a completely new design.
08-30-2010, 05:24 AM
Jowy Atreides

Quote:

Originally Posted by LightSpeed

But the current drivers are not introductory drivers. Fermi already received a performance boost with the latest drivers. Also, why are you ignoring the fact the cayman is a new architecture as well, that can be tweaked for more performance with newer drivers too :shrug:

Don't forget all the "ATI bad drivers hurrrrrrr" comments.

If nvidia drivers are unoptimized and inefficient for a whole year and carvidia himself said it.
Yup, rub in face time
08-30-2010, 05:40 AM
Manicdan

i feel like carfax is going to cause another thread to be locked

my opinion is that nvidia is not a graphics card company anymore
they are HPC
Physx
and drivers

they simply buy games with TWIMTBP, and force you to use their products to enjoy it

i think it would have been easier and cheaper to just make a good graphics card from the start, instead of having to expand your business just to keep your customers you already had.

its time they get back to their roots and build something for gamers
08-30-2010, 05:44 AM
Frontl1ne

Quote:

Originally Posted by Manicdan

its time they get back to their roots and build something for gamers

Didn't you know? The future is CUDA. Who cares about gamers when our graphics cards have 512 CUDA CORES!!!! :clap:
08-30-2010, 05:56 AM
Carfax

Quote:

Originally Posted by Manicdan

i feel like carfax is going to cause another thread to be locked

And how is that? I've been nothing but civil in this thread.. It's not my fault if other people get bent out of shape :rolleyes:

Quote:

its time they get back to their roots and build something for gamers

I don't know what you're talking about. 480 SLi is the fastest gaming setup I've ever owned, and along with 480 Tri SLi, it is the only setup that can comfortably handle the XD resolutions ie 2560x1600 and above.

Ironically, your pure gaming chips stumbles miserably when it comes to the extreme gaming :rolleyes:
08-30-2010, 05:58 AM
Mechanical Man

Quote:

Originally Posted by Frontl1ne

Didn't you know? The future is CUDA. Who cares about gamers when our graphics cards have 512 CUDA CORES!!!! :clap:

Cut down to 480 CUDA cores by design...

:rofl:
08-30-2010, 06:54 AM
LightSpeed

Quote:

Originally Posted by Carfax

I wouldn't say I was ignoring it. It's just that even though Cayman is a new architecture, it will still share very many similarities to it's predecessors.

Since they migrated towards a unified shader architecture, ATI's strategy has been to simply double all the resources on new designs. I think the only thing that ATI didn't double with the R800 was the memory interface, which remained at 256 bit.

And since Cayman won't be getting a die shrink, it seems like it will be a more refined version of the R800, rather than a next generation architecture.

Anyway, driver optimization is an ongoing affair with no end in sight. Cayman will undoubtedly receive performance improvements via driver tweaks, but I doubt it will get as much as Fermi, since Fermi was a completely new design.

Yes, Cayman will be a more refined version of the R800. Think RV670 > RV770. Along with increasing the resources, efficiency was increased as well.

Also, since Fermi already got a performance boost with the new driver, and Cayman isnt a completely new architecture, Id say performance boost would be similar between the two with new drivers. Which leaves Cayman with a higher performance and lower power consumption as well (most likely)

The Nvidia exclusive features dont matter enough to the majority of people to warrant a card thats slower and consumes more power IMO, but to those who think the features are that important, thats perfectly fine. To each, their own
08-30-2010, 07:00 AM
Manicdan

Quote:

Originally Posted by Carfax

And how is that? I've been nothing but civil in this thread.. It's not my fault if other people get bent out of shape :rolleyes:

I don't know what you're talking about. 480 SLi is the fastest gaming setup I've ever owned, and along with 480 Tri SLi, it is the only setup that can comfortably handle the XD resolutions ie 2560x1600 and above.

Ironically, your pure gaming chips stumbles miserably when it comes to the extreme gaming :rolleyes:

your in here trying to preach about how your gaming machine is the fastest in the world

the discussion started because we say that perf per watt and perf per mm2 is way in favor with AMD vs the gf100, which is a fact.

phsyx being awesome, is an opinion
nvidias driver team is going to get you 15% more perf in the next year, is an assumption
nvidia moving to cuda cores rather than gaming perf will earn them more money in the long run, is an assumption.

and all 3 of those things have nothing to do with rumors for the 6000 series cards, but instead push people into an argument very similar to those in the last physx thread that was closed

so please either offer something useful without starting up a my card is bigger than your card war
08-30-2010, 07:14 AM
flopper

Quote:

Originally Posted by Carfax

Unless you're a semiconductor engineer, how could you possibly know that Fermi is power hungry due to "bad design?"

.

are you one?
If you aint how would you even consider my answer to be appropiate as you wouldnt understand it?;)

Its called 460 since Nvidia went over the top with their first generation Fermi.
its bad design, simple as that.

amd on the other hand, well...their 3 last designs and implementations seems to be flawless. The jury still out for the 6000 series.
08-30-2010, 07:17 AM
Calmatory

Quote:

Originally Posted by Carfax

Ironically, your pure gaming chips stumbles miserably when it comes to the extreme gaming :rolleyes:

Extreme gaming. :ROTF:

Honestly. :(
08-30-2010, 07:28 AM
Carfax

Quote:

Originally Posted by Manicdan

your in here trying to preach about how your gaming machine is the fastest in the world

Talk about exaggerating. You guys were the ones bragging as to how the Cayman would own Fermi and blah blah blah. I merely advised to be more skeptical of the benches, and as it turned out, they are likely fake.

Quote:

the discussion started because we say that perf per watt and perf per mm2 is way in favor with AMD vs the gf100, which is a fact.

It's only a fact if you hold to your ignorant view that Fermi is a dedicated gaming processor, when the "fact" is, it's not.

Fermi is much more than just a gaming GPU, and as such, you can't hold it to the same standard as you do Cypress.

Quote:

phsyx being awesome, is an opinion

I agree and even said so. PhysX is an extra optional feature, that not everyone will care for.

Quote:

nvidias driver team is going to get you 15% more perf in the next year, is an assumption

It's an assumption based on recurrent past experience, so it's not as though it's groundless.

Quote:

nvidia moving to cuda cores rather than gaming perf will earn them more money in the long run, is an assumption.

You act as though every assumption is bad, or has no element of truth. :rolleyes:

It's a FACT that the HPC market is more profitable than the gamer market.

As I already mentioned previously, a Quadro video card can cost as much as 5,000 USD! With that kind of price tag, is it a groundless assumption to think that Nvidia investing more resources in HPC won't earn them more money?

You think Nvidia is stupid? Why do you think they designed Fermi in the first place?

Quote:

and all 3 of those things have nothing to do with rumors for the 6000 series cards, but instead push people into an argument very similar to those in the last physx thread that was closed

Seriously, take your blinders off. You act as though you guys have been discussing nothing but 6000 rumors, when in reality, you've been comparing it to Fermi and slamming Fermi while touting the 6000 series as the greatest thing ever when the specs aren't even known! :rolleyes:

Quote:

so please either offer something useful without starting up a my card is bigger than your card war

I'll do that the moment you guys stop acting as though the 6000 series is going to own Fermi.
08-30-2010, 07:38 AM
Calmatory

I do think Nvidia is stupid. If they weren't, Fermi wouldn't have been a failure for them. It didn't meet their expectations, hence why it was failure for them. Saaya's sig sums it up.
08-30-2010, 07:43 AM
Carfax

Quote:

Originally Posted by LightSpeed

Also, since Fermi already got a performance boost with the new driver, and Cayman isnt a completely new architecture, Id say performance boost would be similar between the two with new drivers. Which leaves Cayman with a higher performance and lower power consumption as well (most likely)

Cayman could very likely outperform the 480, while having lower power consumption I agree. I guess we'll just have to wait and see..

Quote:

The Nvidia exclusive features dont matter enough to the majority of people to warrant a card thats slower and consumes more power IMO, but to those who think the features are that important, thats perfectly fine. To each, their own

Well there's more to it than that. A lot of people go with Nvidia because of their driver support.

Whether you agree with that or not, you have to admit that there's a perception that ATI's drivers are inferior to Nvidia's.

As long as I've been a computer enthusiast (which is slightly more than a decade now), that stereotype has persisted.

Quote:

Originally Posted by flopper

are you one?
If you aint how would you even consider my answer to be appropiate as you wouldnt understand it?;)

I'm not, but you don't see me making comments as to whether Fermi had a bad design or not.

That kind of question warrants indepth knowledge that I don't possess.

Quote:

Its called 460 since Nvidia went over the top with their first generation Fermi.
its bad design, simple as that.

Fermi was designed to tap into BOTH the HPC and gaming market. As such, how can you say it's a bad design when Fermi is leading the 5000 series in gaming, and utterly destroying ATI's FirePro series in the professional market.

Quote:

amd on the other hand, well...their 3 last designs and implementations seems to be flawless. The jury still out for the 6000 series.

So flawless they need to use "brilinear" filtering :rolleyes:
08-30-2010, 07:45 AM
Carfax

Quote:

Originally Posted by Calmatory

I do think Nvidia is stupid. If they weren't, Fermi wouldn't have been a failure for them. It didn't meet their expectations, hence why it was failure for them. Saaya's sig sums it up.

Yep, Fermi is a failure when it's leading the 5000 series in most gaming benchmarks (especially DX11), and destroying the FirePro series in professional applications.. :confused::rolleyes:

Not to mention xtreme high resolution gaming, and 3D gaming where again, Nvidia dominates ATI.
08-30-2010, 07:47 AM
Iconyu

Quote:

Originally Posted by Calmatory

I do think Nvidia is stupid. If they weren't, Fermi wouldn't have been a failure for them. It didn't meet their expectations, hence why it was failure for them. Saaya's sig sums it up.

Actually, I think nvidia didn't want to use Fermi for the gaming market. It's powerful but it's not a gaming card really, it's just got so much grunt it can pretend that it is.

Charlie said they had planned a different card but scrapped it and only had Fermi in reserve so they had to make it fit. And I agree with him, Fermi could not really be nvidia's first choice for the gaming market, they would have known it would be a bastard to make but when you're planning to sell at 5k each it's not a issue.
08-30-2010, 07:51 AM
LightSpeed

Quote:

Originally Posted by Carfax

Cayman could very likely outperform the 480, while having lower power consumption I agree. I guess we'll just have to wait and see..

Well there's more to it than that. A lot of people go with Nvidia because of their driver support.

I have to agree there. AMD's driver team seriously needs to get their act together, especially the stuff that get broke after being fixed before.
08-30-2010, 08:17 AM
Carfax

Quote:

Originally Posted by Iconyu

Charlie said they had planned a different card but scrapped it and only had Fermi in reserve so they had to make it fit. And I agree with him, Fermi could not really be nvidia's first choice for the gaming market, they would have known it would be a bastard to make but when you're planning to sell at 5k each it's not a issue.

I think Charlie was talking out of his a$$ if he said that.

Fermi follows the same pattern as previous Nvidia GPUs, in that the Quadro series is practically the same as the consumer GPU, except for different drivers, BIOS and a lot more VRAM.

Quote:

Originally Posted by LightSpeed

I have to agree there. AMD's driver team seriously needs to get their act together, especially the stuff that get broke after being fixed before.

You said it, not me :p:
08-30-2010, 08:23 AM
Calmatory

Quote:

Originally Posted by Carfax

Yep, Fermi is a failure when it's leading the 5000 series in most gaming benchmarks (especially DX11), and destroying the FirePro series in professional applications.. :confused::rolleyes:

Not to mention xtreme high resolution gaming, and 3D gaming where again, Nvidia dominates ATI.

So you're talking from consumers point of view now? Interesting. Stick to that from now on. :)

It took Nvidia 6 months of lies(all that "THIS IS FERMI!!" bull:banana::banana::banana::banana:) and very lame marketting(Announcement of an announcement. What?) to get it out. And when it got out... power consumption and crippled core. Nvidia failed. All you really need to do is to take a look at DX11 cards market share to see this. Nvidia is under 10 %. Definitely not something they aimed for, right?
08-30-2010, 09:18 AM
LightSpeed

Quote:

Originally Posted by Carfax

Yep, Fermi is a failure when it's leading the 5000 series in most gaming benchmarks (especially DX11), and destroying the FirePro series in professional applications..

Not to mention xtreme high resolution gaming, and 3D gaming where again, Nvidia dominates ATI

In xtreme high resolution gaming, I wouldnt say Nvidia dominates ATI, but leads them. That is, at the expense of massive power consumption.

Dominates is a strong word :)

Quote:

Originally Posted by Carfax

You said it, not me :p:

Yeah, I forgot to add the driver part. But still, the conclusion i made is unchanged. Nv features + better driver is outweighed by a faster + less power hungry card to most people.

From what I hear, the ATI driver team has no excuse to be poor, they have enough people and resources. I gotta say they do bring about minor performance boosts recently and have started to fix stuff more frequently => general improvement. But why break stuff they already fix? It completely baffles me :shrug:
08-30-2010, 09:33 AM
Sn0wm@n

LOL why is every amd related thread getting attention from tards ??? :(
08-30-2010, 09:54 AM
Jowy Atreides

Quote:

Originally Posted by Sn0wm@n

LOL why is every amd related thread getting attention from tards ??? :(

Fear. People's shares are at stake.
08-30-2010, 10:01 AM
cegras

What? In terms of recent market share according to the Steam survey, ATI clearly leads nvidia in DX10 and 11 GPUs. All of nvidia's large marketshare is legacy left over from the 7xxx series.
08-30-2010, 10:15 AM
Jowy Atreides

Quote:

Originally Posted by cegras

What? In terms of recent market share according to the Steam survey, ATI clearly leads nvidia in DX10 and 11 GPUs. All of nvidia's large marketshare is legacy left over from the 7xxx series.

Not market share, stock value dude.

You think there aren't people with chip stock who aren't aware that speculation and malicious rumours can affect the value of their assets?

When fake benchmarks appear, don't assume it's an attention whore. It could be anti competition fixing.
08-30-2010, 10:20 AM
informal

Quote:

Originally Posted by Jowy Atreides

Fear. People's shares are at stake.

That's unfortunately true.
08-30-2010, 02:57 PM
Carfax

Quote:

Originally Posted by Calmatory

Nvidia is under 10 %. Definitely not something they aimed for, right?

I would think it obvious that this is because ATI had a 6 month head start on Nvidia :shakes:

Quote:

Originally Posted by LightSpeed

In xtreme high resolution gaming, I wouldnt say Nvidia dominates ATI, but leads them. That is, at the expense of massive power consumption.

Dominates is a strong word :)

Dominates is a strong word, but it's the truth.

Read this review from HardOCP about Nvidia surround vs Eyefinity using 480 SLi and HD 5870 2GB Crossfire @ 5760x1200

In the words of Kyle Bennett:

Quote:

It is disturbingly obvious in our testing that GeForce GTX 480 SLI is superior to Radeon HD 5870 Eyefinity6 2GB and Radeon HD 5870 1GB CrossFireX configurations in 3x1 multi-display gaming. GeForce GTX 480 SLI is a powerhouse of performance at high resolution multi-display gaming. In every game we played it didn’t just come out on top a little, it completely owned the two HD 5870 CrossFireX setups.

Quote:

Yeah, I forgot to add the driver part. But still, the conclusion i made is unchanged. Nv features + better driver is outweighed by a faster + less power hungry card to most people.

I would have to disagree with you on this. A fast card is nothing without good drivers that exploit the hardware to the fullest possible degree.

On that note, recent Nvidia drivers could do with some work as well. The 200xx drivers are very immature and should not have been released so early if you ask me.

Only with the beta release of the 259.32 drivers have I noticed that they are finally starting to get great optimization.

Quote:

From what I hear, the ATI driver team has no excuse to be poor, they have enough people and resources. I gotta say they do bring about minor performance boosts recently and have started to fix stuff more frequently => general improvement. But why break stuff they already fix? It completely baffles me :shrug:

Yep, like I said, as long as I've been a computer enthusiast, the notion that ATI has poor drivers has been around.

Thats more than 10 years now, so thats saying something!

Quote:

Originally Posted by Sn0wm@n

LOL why is every amd related thread getting attention from tards ??? :(

You mean tards that jump at every piece of rumor and treat it as though it's fact? :ROTF:
08-30-2010, 03:10 PM
spursindonesia

Or perhaps a tard that can't stay on topic and running amok spreading FUDs in a frenzy mode & crapping the thread all over the place with mine card is bigger than yours attitude ? :ROTF:
08-30-2010, 03:17 PM
Carfax

Quote:

Originally Posted by spursindonesia

Or perhaps a tard that can't stay on topic and running amok spreading FUDs in a frenzy mode & crapping the thread all over the place with mine card is bigger than yours attitude ? :ROTF:

LOL yeah, as if you guys were staying on topic :rofl:

How is,"Fermi a failure" on topic?

Or,"Nvidia is way behind on perf per mm2."

And in your own words,"I think GTX 480 will still be SUPERIOR .... as a room heater, remember, the cold days are arriving in the northern hemisphere."

Yeah, very on topic :down:
08-30-2010, 03:18 PM
kadozer

Carfax did you miss this article from H http://www.hardocp.com/article/2010/...8a_performance Would be interesting to test them out now. AMD's driver team is behind Nvidia's and needs to be more on top of fixing problems faster, is 1 thing I do agree with.

Just wait till the 6 series, after all you are comparing Fermi Nvidia's current gen to AMD's old one.

I'd like to see the next Fermi coming out this Q4 vs the 6 series.
08-30-2010, 03:32 PM
Carfax

Quote:

Originally Posted by kadozer

Carfax did you miss this article from H http://www.hardocp.com/article/2010/...8a_performance Would be interesting to test them out now. AMD's driver team is behind Nvidia's and needs to be more on top of fixing problems faster, is 1 thing I do agree with.

Yeah I saw that.. But, showing skepticism towards "miracle drivers" is a good thing.

In fact, a HardOCP forum member noted that Vsync and AA did not appear to be working properly with the new drivers

More.

Quote:

Just wait till the 6 series, after all you are comparing Fermi Nvidia's current gen to AMD's old one.

Both Cypress/Hemlock and 480/470 are ATI and NVidia's first generation DX11 cards, so I don't think the comparison is unjustified.
08-30-2010, 03:41 PM
kadozer

Quote:

Originally Posted by Carfax

Yeah I saw that.. But, showing skepticism towards "miracle drivers" is a good thing.

In fact, a HardOCP forum member noted that Vsync and AA did not appear to be working properly with the new drivers

More.

Both Cypress/Hemlock and 480/470 are ATI and NVidia's first generation DX11 cards, so I don't think the comparison is unjustified.

6 months late is not justified. That is a good spin though. The miracle drivers bit is also good. I don't think anybody from AMD claimed them to be that good.
08-30-2010, 03:50 PM
Eastcoasthandle

Quote:

Originally Posted by Carfax

Yeah I saw that.. But, showing skepticism towards "miracle drivers" is a good thing.

In fact, a HardOCP forum member noted that Vsync and AA did not appear to be working properly with the new drivers

More.

Both Cypress/Hemlock and 480/470 are ATI and NVidia's first generation DX11 cards, so I don't think the comparison is unjustified.

There is nothing wrong with AA on Radeon cards as far as I know. However reports of it's use in game is not uncommon in BC2. Which indicate that it's not specific to any particular driver or video card. As for Vsync as far as I can see it works fine in BC2 in both single and multi player. So I honestly don't foresee any problem with the next gen cards.
08-30-2010, 05:12 PM
saaya

Quote:

Originally Posted by Carfax

Unless you're a semiconductor engineer, how could you possibly know that Fermi is power hungry due to "bad design?"

oh, your saying its designed to run hot on purpose then?
but why, you must be an ee engineer to make a claim like that! :lol:

Quote:

Originally Posted by Carfax

I don't think most people here understand that Fermi is the first TRULY computational GPU. All those things you said about convergent technologies into one neat package, are made manifest in Fermi....in the sense that Fermi has many CPU like qualities such as being fully IEEE compliant, support for C++, having dedicated caches and support for much greater VRAM.

oh i think everybody knows that... we all saw those dozens of marketing slides from nvidia of how great fermi is and all the things you can do with it... in theory... i just think nobody cares...
thats not the reason somebody spends 250-500$ on a VGA = video graphics accelerator these days... when they buy a vga, guess what, they want it to ACCELERATE GRAPHICS! shocker! :eek:

you can sell chainsaw as a knife and advertise it as "can cut down a tree", itll still suck as a knife and nobody will want to carry such a huge expensive thing from the kitchen to the breakfast table every morning... :D
and all your talk about how it also serves as an "urban violence solver" and "stress relief super fun toy" wont make it a better knife... :D

Quote:

Originally Posted by Carfax

The very fact that a GPU like Fermi, which isn't even a pure gamer chip can outperform Cypress (which is dedicated towards gaming) speaks volumes in and of itself.

and a rocket outperforms a ferrari, does that make it a better race car? :idea:

Quote:

Originally Posted by Carfax

It wouldn't surprise me if Fermi ends up being 20% faster than it was on introduction, simply due to driver optimizations.

on average? hah! hahahahahhaha!

Quote:

Originally Posted by Carfax

It's only a fact if you hold to your ignorant view that Fermi is a dedicated gaming processor, when the "fact" is, it's not.

can you stop jumping between these two points please? your making my head spin!

fermi isnt a graphics processor!
fermi is an awesome graphics processor!
fermi isnt a graphics processor!

:confused2

Quote:

Originally Posted by Carfax

Fermi is much more than just a gaming GPU, and as such, you can't hold it to the same standard as you do Cypress.

again!

Quote:

Originally Posted by Carfax

It's an assumption based on recurrent past experience, so it's not as though it's groundless.

show us... ive seen several articles on the matter and they all proved that driver improvements of various gpus in the past years have resulted in a ~10% boost at most...
so which past experience are you talking about?
which past gpu makes youy think expecting a 20% boost is reasonable for fermi?
check out the first reviews of rv870 and compare them to the latest reviews comparing it to gf104 etc... and youll see that rv870 saw a boost of less than 10% throughout its lifetime...

Quote:

Originally Posted by Carfax

It's a FACT that the HPC market is more profitable than the gamer market.

it is? :confused:

Quote:

Originally Posted by Carfax

As I already mentioned previously, a Quadro video card can cost as much as 5,000 USD! With that kind of price tag, is it a groundless assumption to think that Nvidia investing more resources in HPC won't earn them more money?

quadro cards are aimed at hpc? :confused: i thought it was tesla?

Quote:

Originally Posted by Carfax

You think Nvidia is stupid? Why do you think they designed Fermi in the first place?

at that time they were ahead in the graphics field by quite a fair bit, and they probably thought they could get away with all the gpgpu baggage on a gpu... do you think if given the chance nvidia would do the same thing again? i dont think so...

Quote:

Originally Posted by Carfax

I'll do that the moment you guys stop acting as though the 6000 series is going to own Fermi.

you think its unreasonable to expect a new part to outperform an old part? :eh:

Quote:

Originally Posted by Iconyu

Charlie said they had planned a different card but scrapped it and only had Fermi in reserve so they had to make it fit. And I agree with him, Fermi could not really be nvidia's first choice for the gaming market, they would have known it would be a bastard to make but when you're planning to sell at 5k each it's not a issue.

hmmm where did he say that? do you remember the name of the article?

Quote:

Originally Posted by LightSpeed

I have to agree there. AMD's driver team seriously needs to get their act together, especially the stuff that get broke after being fixed before.

yeah... i agree...
some of the things ive heard in the past couple of months made me glad im on nvidia cards atm... while there are annoying things in their drivers as well, its nothing major...
08-30-2010, 06:27 PM
Iconyu

Man, he wrote so much it's hard to find something he said only once or twice.

Quote:

How did this come about? Sources in Santa Clara tell SemiAccurate that GF100 was never meant to be a graphics chip, it started life as a GPGPU compute chip and then abandoned. When the successor to the G200 didn't pan out, the GF100 architecture was pulled off the shelf and tarted up to be a GPU. This is very similar to what happened to ATI's R500 Xenos, but that one seems to have worked out nicely in the end.

http://www.semiaccurate.com/2010/01/...anufacturable/

This rant is pretty famous, you'll probably remember it. And after reading it through, it actually is "semi-accurate". While he did overplay how bad the 480GTX would be, he was spot on about the card being "unmanufacturable", nvidia have not been able to market a 512 shader part after a year of making them. They are either broken or they can't get to target clocks without going over 300 watts.

ATI have much more room to move on 40nm, beating the 480 isn't impossible for them as long as they're careful about it.
08-30-2010, 07:08 PM
kadozer

Quote:

Originally Posted by Iconyu

Man, he wrote so much it's hard to find something he said only once or twice.

http://www.semiaccurate.com/2010/01/...anufacturable/

This rant is pretty famous, you'll probably remember it. And after reading it through, it actually is "semi-accurate". While he did overplay how bad the 480GTX would be, he was spot on about the card being "unmanufacturable", nvidia have not been able to market a 512 shader part after a year of making them. They are either broken or they can't get to target clocks without going over 300 watts.

ATI have much more room to move on 40nm, beating the 480 isn't impossible for them as long as they're careful about it.

Its pretty amusing actually. Guy is talking like the pre release days of Fermi re-living its glorious pre hype release talk. Cuda cores, computational power, etc :up: We needed a fun thread its been awhile.

AMD is pretty quiet about this upcoming series. That's always a good thing unlike Nvidia that talked up Fermi and we all know how that turned out. For a next gen gpu = meh.
08-30-2010, 07:34 PM
spursindonesia

Quote:

Originally Posted by Carfax

LOL yeah, as if you guys were staying on topic :rofl:

How is,"Fermi a failure" on topic?

Or,"Nvidia is way behind on perf per mm2."

And in your own words,"I think GTX 480 will still be SUPERIOR .... as a room heater, remember, the cold days are arriving in the northern hemisphere."

Yeah, very on topic :down:

LOL, atleast i start ALL of my pretenses with "IF" word, i don't command any claim or knowledge that they're all and nothing but the truth. :ROTF: OTOH, how are your posts that claim nVidia's supremacy FUDs regarding many things constructive to the discussion in this thread ? If your rebuttal is concentrated on making the point of the falsity of the leaked scores, then make it so, i have no problem with that. Heck, even claiming for sure that these new graphic chips from ATi would suck, it won't annoy me at all, you will be held accountable for anything you've said today by the forum, not me. :clap:

I'm so puzzled with cult like attitude by some forum members regarding products/companies, it's getting ridiculously ideological & pathetic. I've always get the best bang for the bucks products, manufacturer be damned. I've had good times with both nVidia & ATi cards while they're at the top of the game, such as GeForce 3 Ti, 6600 GT, 8800 GT for the green camp, Radeon 9500, X850 Pro, X1900 XT 256, and the current HD 4870. If i have the need for a new card at the moment, i might easily buy a GTX 460 1 GB for its perceived & appreciated great value. This is just so weird and sad IMHO. :cool:

EDIT: To make my self clear, my comment on "I think GTX 480 will still be SUPERIOR .... as a room heater, remember, the cold days are arriving in the northern hemisphere." was forwarded after you make all those FUDs carpet bomb claims. When you ROFLMAOing the leaks and people who hope for the best, i keep my silence.

Do i think Fermi has advantage in HPC arena ? Resounding YES. Do i think nVidia offers some exclusive features ? True, but perception of people regarding its importance & usefulness might vary. And regarding driver support + game IQ, i and many people that i know feel indifferent regarding these cause both camps have their up & down of each own, nothing to get bittered about.

In the end, we're talking about these products as Video Graphic Accelerator, and when the new generation of one company pulls ahead of the older generation of another, i call that a good technological advance and moving on, nothing to get defensive about like what you've done all over this thread. IF the numbers are true, then GTX 480 would be inferior as a graphic chip, aka slower for gaming, nothing shame about that since it has happened for eternity, and you may keep enjoying your personal preferences no problem. Just keep your FUD opinions for your self, else it will be called accordingly by others.
08-31-2010, 12:21 AM
mindfury

Oops...
08-31-2010, 12:35 AM
cold2010

Quote:

Originally Posted by mindfury

ASUS HD6870 ROG video:
http://v.youku.com/v_show/id_XMjAyNjAyODA4.html

http://v.youku.com/v_show/id_XMjAyNjAzMDg4.html

:rofl:

Quote:

This is just a concept design i made in 3ds max 2009. I redesigned the intake fan with dual turbine fans. This model takes two PCI-E 8 pins for great overclocking. The card is approx. 11.5 inches long. Rendered with V-ray. Hope you enjoy.

http://www.youtube.com/watch?v=AS7YqRpvZ7Y
08-31-2010, 12:35 AM
tajoh111

Quote:

Originally Posted by mindfury

ASUS HD6870 ROG video:
http://v.youku.com/v_show/id_XMjAyNjAyODA4.html

http://v.youku.com/v_show/id_XMjAyNjAzMDg4.html

Considering the floating letters and everything, I am pretty sure that is a render.
08-31-2010, 12:39 AM
flopper

Quote:

Originally Posted by mindfury

ASUS HD6870 ROG video:
http://v.youku.com/v_show/id_XMjAyNjAyODA4.html

http://v.youku.com/v_show/id_XMjAyNjAzMDg4.html

black = sexy.
08-31-2010, 01:23 AM
saaya

Quote:

Originally Posted by Iconyu

Man, he wrote so much it's hard to find something he said only once or twice.
http://www.semiaccurate.com/2010/01/...anufacturable/

This rant is pretty famous, you'll probably remember it. And after reading it through, it actually is "semi-accurate". While he did overplay how bad the 480GTX would be, he was spot on about the card being "unmanufacturable", nvidia have not been able to market a 512 shader part after a year of making them. They are either broken or they can't get to target clocks without going over 300 watts.

ATI have much more room to move on 40nm, beating the 480 isn't impossible for them as long as they're careful about it.

you did it nevertheless! :D
thx man, appreciated :toast:

about fermi never being meant as a graphics card... i dont think so... its limited enough as a gp processor to make it unlikely that it was originally built for that alone... but i wouldnt be surprised if gf100 was originally planned for 32nm... but since they were ahead of ati by so much (back then) and they knew about intel lrb... they probably thought they could use a shortcut and squeeze it into 40nm... very risky... and it kinda blew up in their face, heh :D

if it would have worked out... very good... but really, it was pretty weird to expect it to work out fine... there were so many things that could go wrong, and its soooo unlikely it all would have worked out...

but you gotta give props to nvidia, for how much went wrong, they took it amazingly well imo... they took a big blow but they didnt even get on their knees let alone fall... but from now on nvidia needs to make stable steps, if they get another serious hit they will def be on their knees soon...

Quote:

Originally Posted by spursindonesia

LOL, atleast i start ALL of my pretenses with "IF" word, i don't command any claim or knowledge that they're all and nothing but the truth.

:toast:
if, i think, i believe and imo are important catalysts in online discussions :D
... in my opinion ^^

Quote:

Originally Posted by mindfury

ASUS HD6870 ROG video:
http://v.youku.com/v_show/id_XMjAyNjAyODA4.html

http://v.youku.com/v_show/id_XMjAyNjAzMDg4.html

hmmmmmm
that weird fan system doesnt look very realistic... and why does it say Asus Crosshair at the end? :confused:
08-31-2010, 02:14 AM
mindfury

Quote:

Originally Posted by saaya

that weird fan system doesnt look very realistic... and why does it say Asus Crosshair at the end? :confused:

Because it is fake,I got fooled,damn...
08-31-2010, 04:08 AM
mindfury

http://www.donanimhaber.com/ekran-ka...e-detaylar.htm

Radeon HD 6700 series will launch on October 12.
08-31-2010, 04:30 AM
blindbox

You're kidding! 1 month and 12 days away ! :shocked:
08-31-2010, 05:14 AM
LightSpeed

IIRC the 67xx isnt the new architecture, its a slightly modified Cypress with different clocks/PCB with some SP's disabled most probably

I might be wrong, but I do trust the source.
08-31-2010, 05:30 AM
Calmatory

So renaming.. :( AMD FAIL :down:
08-31-2010, 05:33 AM
~CS~

Quote:

Originally Posted by Calmatory

So renaming.. :( AMD FAIL :down:

And how did you come to that conclusion ? Or you are just flame bating ? ..if it its the latter nvm.
08-31-2010, 05:36 AM
madcho

i don't think this is renaming.
08-31-2010, 05:36 AM
mindfury

Quote:

Originally Posted by LightSpeed

IIRC the 67xx isnt the new architecture, its a slightly modified Cypress with different clocks/PCB with some SP's disabled most probably

I might be wrong, but I do trust the source.

6700 series codename is Barts,which is member of NI family.So it has to be the new architecture.
08-31-2010, 05:59 AM
Sn0wm@n

LOL this isnt renaming .... ati wont do it when they are barely dominating the market .... its easy to see that this new card will have noticeable improvement
08-31-2010, 06:15 AM
ajaidev

HD6000 - first ATI card with digital PWM:

http://translate.google.com/translat...e&tl=en&swap=1
08-31-2010, 06:26 AM
Calmatory

Quote:

Originally Posted by Sn0wm@n

LOL this isnt renaming .... ati wont do it when they are barely dominating the market .... its easy to see that this new card will have noticeable improvement

Renaming also gives some benefits. Redo cypress with slight changes and different clocks as Barts and slap it out as HD6700 series. One could argue it isn't renaming, while other could argue it is. The fact is, that for minimal effort they can resell the old, albeit modified hardware for almost the same margins(Only now, since the GPU prices are inflated anyway, thanks Nvidia. :p:).

Now read this before you reply: I am not claiming that the kind of renaming I just describes would really match the renaming Nvidia is doing. I am saying that it is actually BETTER for a company to recycle old stuff, since renaming would make sense.

If the 4+1 -> 4 rumour is true, then obviously renaming wouldn't make any sense at all. :p:
08-31-2010, 06:38 AM
Sn0wm@n

we dont have the same deffinition of renaming then .... amd had to do it that way ... they were forced to take an old design improve it and use the same 40nm node ... if they really did some good improvement then i wouldnt consider it a rename ... if they barely did some tweaks its a card that we could live without ..
08-31-2010, 06:53 AM
Calmatory

Renaming makes no sense though, so it's hardly a possibility then. :p:
08-31-2010, 06:55 AM
mindfury

http://www.chiphell.com/thread-120059-1-1.html

Quote:

比HD5870稍长一些(很少一点点);
频率现在是850MHz,但不代表最终零售不变.
TDP在200W以内.
散热器使用了均热板,很给力.(异形,并不是平面的,估计又是高级货)
nApoleon 发表于 2010-8-31 22:50

Cayman pro will a bit longer than HD5870.
Core clcok=850Mhz
TDP<200W
Vapor heatsink(Not flat)
08-31-2010, 07:36 AM
Calmatory

What exactly is that vapor heatsink? Something new?

IIRC at least Nordic Hardware was touting about the new cooling technology of having a vapor chamber in the bottom of the heatsink for R800 products back in summer of 2008. Afaik, that wasn't the case... Maybe now?
08-31-2010, 07:37 AM
Shadov

Quote:

Originally Posted by Calmatory

Renaming makes no sense though, so it's hardly a possibility then. :p:

If it has improved performance and tessellation would you call it renaming? :cool:
08-31-2010, 07:46 AM
~CS~

Quote:

Originally Posted by Calmatory

What exactly is that vapor heatsink? Something new?

IIRC at least Nordic Hardware was touting about the new cooling technology of having a vapor chamber in the bottom of the heatsink for R800 products back in summer of 2008. Afaik, that wasn't the case... Maybe now?

Actually it is the case,all reference 5970 cards use a vapor chamber.It's a basic tech similar to a heat pipe but more surface area.
08-31-2010, 08:32 AM
blindbox

Quote:

Originally Posted by Calmatory

What exactly is that vapor heatsink? Something new?

IIRC at least Nordic Hardware was touting about the new cooling technology of having a vapor chamber in the bottom of the heatsink for R800 products back in summer of 2008. Afaik, that wasn't the case... Maybe now?

Sapphire has been using vapor heatsink (their Toxic products, and now their Vapor-X'es) since the Radeon HD 3xxx. AMD used it for 5970 reference design.

Quote:

Originally Posted by LightSpeed

IIRC the 67xx isnt the new architecture, its a slightly modified Cypress with different clocks/PCB with some SP's disabled most probably

I might be wrong, but I do trust the source.

I believe that's some pretty old info.
08-31-2010, 08:43 AM
spursindonesia

Vapor chamber, basically the whole base of the heatsink acts as a giant flat heatpipe, sort of like that, correct ? :shrug:

Regarding the renaming issue, i dunno man, i don't think it's wise and purposeful for ATi going this avenue when Evergreen family is still more than decent against competition, it will simply erode the goodwill from its costumers. It will be such a shame for them if with all the time afforded with Fermi's debacle, they just sit doing nothing and resorting to renaming & milking the market tactic. :down:
08-31-2010, 09:28 AM
LightSpeed

Quote:

Originally Posted by blindbox

I believe that's some pretty old info.

Not really, I heard it from someone in the know only a few days ago.

Im not saying they are renaming, but that the 67xx will be a modified Cypress and the 68xx will be a new architecture.

Time will certainly tell :)
08-31-2010, 09:41 AM
blindbox

Quote:

Originally Posted by LightSpeed

Not really, I heard it from someone in the know only a few days ago.

Im not saying they are renaming, but that the 67xx will be a modified Cypress and the 68xx will be a new architecture.

Time will certainly tell :)

Well I heard it months ago :confused:

NI is touted to be a mix-match architecture, taking points from both Evergreen and SI. Plus, the naming is already out, indicating NI. I'm just taking chances from AMD's track record. It won't be a renaming.

Time always tells :up:
08-31-2010, 09:43 AM
Mysterfix

Quote:

Originally Posted by spursindonesia

Vapor chamber, basically the whole base of the heatsink acts as a giant flat heatpipe, sort of like that, correct ? :shrug:

Regarding the renaming issue, i dunno man, i don't think it's wise and purposeful for ATi going this avenue when Evergreen family is still more than decent against competition, it will simply erode the goodwill from its costumers. It will be such a shame for them if with all the time afforded with Fermi's debacle, they just sit doing nothing and resorting to renaming & milking the market tactic. :down:

I wouldn't say they are sitting on their hands but you can only do so much on the current node. They already had new tech ready to go on the smaller nodes but manufacturing was not able to deliver so now they fall on to their backup plan in order to be able to offer something fresh for sale. Most people won't know and won't care if these cards modified versions of other tech as long as the performance is there.
08-31-2010, 09:43 AM
bill_d

Quote:

Originally Posted by Calmatory

What exactly is that vapor heatsink? Something new?

IIRC at least Nordic Hardware was touting about the new cooling technology of having a vapor chamber in the bottom of the heatsink for R800 products back in summer of 2008. Afaik, that wasn't the case... Maybe now?

no not new Sapphire
http://www.newegg.com/Product/Produc...-893-_-Product
08-31-2010, 09:51 AM
Calmatory

Ah. The Vapor chamber wasn't used for other than double-GPU 5970, and it is now rumoured to be utilized in at least HD68xx cards? Anyone have any idea of any cooling performance improvement, or is it even just something fancy to market ("200 % BETTER HEATSINK-BASE HEAT CONDUCTIVITY FOR ÜBER PERFORMANCE AND NORTH-POLE LIKE TEMPERATURES!!", with no practical benefits in temps)?
08-31-2010, 10:47 AM
LightSpeed

Quote:

Originally Posted by blindbox

Well I heard it months ago :confused:

NI is touted to be a mix-match architecture, taking points from both Evergreen and SI. Plus, the naming is already out, indicating NI. I'm just taking chances from AMD's track record. It won't be a renaming.

Time always tells :up:

Yeah, lets just wait and watch :up:

I think Cayman will be bigger than Cypress, designed for 28/32 nm, consume more power and perform much faster, and the rest of the lot (Turks, Barts etc) will be modified Cypress cores.

Thats not renaming, and thats not even a bad idea IMO as it saves R & D and the next complete refresh from top to bottom will be SI (but it might not be the name of islands :yepp: )
08-31-2010, 04:56 PM
Carfax

Quote:

Originally Posted by saaya

oh, your saying its designed to run hot on purpose then?
but why, you must be an ee engineer to make a claim like that! :lol:

You obviously don't know jack about Fermi dude.. Fermi runs hotter and consumes more power than Cypress because it has 40% more transistors! :rolleyes:

Jeeze, isn't it obvious? People need to do their research before making silly claims..

And the GTX 480 doesn't even run that hot. At stock clocks, it usually tops out at 92 or 93c with the fan on average, well within the 105c threshold.

The initial reviews exaggerated the temps because they used open air test stands.

Mine are overclocked by 21% and the highest temps I ever see are 77c and 87c respectively.....with the fan on 100% of course :D

Quote:

thats not the reason somebody spends 250-500$ on a VGA = video graphics accelerator these days... when they buy a vga, guess what, they want it to ACCELERATE GRAPHICS! shocker! :eek:

Saaya, are you implying that Fermi doesn't accelerate graphics?

Seriously.... :shakes:

Quote:

you can sell chainsaw as a knife and advertise it as "can cut down a tree", itll still suck as a knife and nobody will want to carry such a huge expensive thing from the kitchen to the breakfast table every morning... :D
and all your talk about how it also serves as an "urban violence solver" and "stress relief super fun toy" wont make it a better knife... :D

Sounds like a Chinese proverb :ROTF:

Quote:

fermi isnt a graphics processor!
fermi is an awesome graphics processor!
fermi isnt a graphics processor!

:confused2

I already made myself quite clear. Fermi is a graphics processor that has vastly increased computational abilities over previous generations.

Quote:

which past gpu makes youy think expecting a 20% boost is reasonable for fermi?

I didn't say it was reasonable, just possible based on the premise that Fermi is such a radically different architecture.

Anyway, Nvidia has pulled it off before. Ask anyone that owned a Geforce 3 about the Detonator 20xx drivers. It increased performance by nearly 20% across the board, and up to 50 and 60% in specific games and benchmarks. Basically, it turned the Geforce 3 (Nvidia's first DX8 part) from an ugly duckling into a Swan.

Quote:

check out the first reviews of rv870 and compare them to the latest reviews comparing it to gf104 etc... and youll see that rv870 saw a boost of less than 10% throughout its lifetime...

As many people have noted (including many ATI owners themselves), ATI is sorely lacking in the driver department compared to Nvidia.

Quote:

quadro cards are aimed at hpc? :confused: i thought it was tesla?

You're right, it's Tesla. Quadro is more workstation oriented.

Quote:

at that time they were ahead in the graphics field by quite a fair bit, and they probably thought they could get away with all the gpgpu baggage on a gpu... do you think if given the chance nvidia would do the same thing again? i dont think so...

The Fermi architecture was always designed to be scaleable, so thats irrelevant.

Quote:

you think its unreasonable to expect a new part to outperform an old part? :eh:

It's not unreasonable, just not guaranteed. And the fact that Cayman will be more of a revision than a new architecture, isn't exactly doing it any favors.

It will probably be faster more than likely, but not nearly as much as some of the fanboys in this thread have suggested.

Where Cayman will win big, is on power usage....
08-31-2010, 05:10 PM
Carfax

Quote:

Originally Posted by spursindonesia

I'm so puzzled with cult like attitude by some forum members regarding products/companies, it's getting ridiculously ideological & pathetic. I've always get the best bang for the bucks products, manufacturer be damned. I've had good times with both nVidia & ATi cards while they're at the top of the game, such as GeForce 3 Ti, 6600 GT, 8800 GT for the green camp, Radeon 9500, X850 Pro, X1900 XT 256, and the current HD 4870. If i have the need for a new card at the moment, i might easily buy a GTX 460 1 GB for its perceived & appreciated great value. This is just so weird and sad IMHO. :cool:

You get points for "appearing" to be contrite :rolleyes:

Quote:

EDIT: To make my self clear, my comment on "I think GTX 480 will still be SUPERIOR .... as a room heater, remember, the cold days are arriving in the northern hemisphere." was forwarded after you make all those FUDs carpet bomb claims. When you ROFLMAOing the leaks and people who hope for the best, i keep my silence.

What FUD comments? :confused:

Quote:

In the end, we're talking about these products as Video Graphic Accelerator, and when the new generation of one company pulls ahead of the older generation of another, i call that a good technological advance and moving on, nothing to get defensive about like what you've done all over this thread. IF the numbers are true, then GTX 480 would be inferior as a graphic chip, aka slower for gaming, nothing shame about that since it has happened for eternity, and you may keep enjoying your personal preferences no problem. Just keep your FUD opinions for your self, else it will be called accordingly by others.

Again, what FUD opinions?
08-31-2010, 05:34 PM
sniper_sung

roadmap from chiphell:

http://img829.imageshack.us/img829/1278/roadmap.jpg
08-31-2010, 05:39 PM
-Sweeper_

Quote:

Originally Posted by sniper_sung

roadmap from chiphell:

384-bit? :confused:
08-31-2010, 05:50 PM
Iconyu

Wouldn't that make all the benchmarks fake? As I certainly can't see Bart's owning a 480GTX
08-31-2010, 06:12 PM
tajoh111

Quote:

Originally Posted by sniper_sung

roadmap from chiphell:

http://img829.imageshack.us/img829/1278/roadmap.jpg

Is that referring to performance or pricing. Because it is pricing which is likely, its looks like everything getting priced to what I suspected it would. A 6770 being priced like a 5830... ouch. I think the 6650 and 6670 have to stay at the pricing they are at because they still have to keep some cards under 100 dollars. Man though, if the 6970 gets priced at 699 or 799. The 6950 is priced like a 5970 and those cards are already 700 dollars. Man thats going to suck for the consumer. Makes me scared how pricey the black editions or limited editions will get. Want to replace the 4870x2 but man its going to be pricey.

If it is in regards to performance which is less likely, I think its pretty underwhelming to see the 6850 slower than a 5870.

It wouldn't surprise both are true and performance is as the chart shows as well as pricing is going up for almost the entire line up.
08-31-2010, 06:29 PM
kadozer

Lol @ trying to derail this thread about the upcoming 6 series. Greeeeen Envy. Nuff said.
08-31-2010, 07:00 PM
demonkevy666

I find it odd that the 6850 would still be under 5870. :/
08-31-2010, 09:18 PM
saaya

Quote:

Originally Posted by sniper_sung

roadmap from chiphell:

http://img829.imageshack.us/img829/1278/roadmap.jpg

hmmm that doesnt sound that good...
more realistic, but a bit dissapointing...
looks like the 6870 will be close to a gtx480, but definitely not faster... might not even match it...

im really dissapointed that barts doesnt reach 5850 performance...
that would have been killer... a 150-200$ 5850...
it looks like the 6000 series will be a mere update, if this positioning is correct.
this looks like it doesnt deserve the 6000 name imo, according to this table it should rather be:

5990
5980

5890
5880
5860

5790
5780
08-31-2010, 09:25 PM
Zloyd

Is this reliable at all ? I mean 6850 to not perform as well as a 5870 ?

That just sounds wrong , well maybe its a little overeaction because it could perform almost as good at a signifiantly lower price and the same with 6870 matching 480 but lower prices.
08-31-2010, 09:27 PM
Jowy Atreides

Quote:

Originally Posted by saaya

hmmm that doesnt sound that good...
more realistic, but a bit dissapointing...
looks like the 6870 will be close to a gtx480, but definitely not faster... might not even match it...

The 5870 in some games where ATI is favoured already bests the 480.
Since SI strengthens the weakest areas of 5 series performance and also includes a more efficient 4 shader model (and more of them) with a higher memory amount AND bandwidth (as well as more tesselation modules) then it's not very likely that the 6870 will be anything less than the 480 in the worst case scenario.

Definitely faster.
08-31-2010, 09:29 PM
tajoh111

Quote:

Originally Posted by saaya

hmmm that doesnt sound that good...
more realistic, but a bit dissapointing...
looks like the 6870 will be close to a gtx480, but definitely not faster... might not even match it...

im really dissapointed that barts doesnt reach 5850 performance...
that would have been killer... a 150-200$ 5850...
it looks like the 6000 series will be a mere update, if this positioning is correct.
this looks like it doesnt deserve the 6000 name imo, according to this table it should rather be:

5990
5980

5890
5880

5790
5780

For some reason, I think this is more about pricing than anything. It would seem too screwed up if the 6850 never caught up to the 5870 when there is only a 15 percent between the 5850 and the 5870. It would have to be like only 10% faster and to call that a 6850 would be shameful and misleading.

It would kind of mirror what NV has done with gtx 285 and gtx 275 those and could simply be a super minor update.

But only 10% difference. This could be fake and all but considering the timing of this lineup and the obvious flaws found in some of the benchmarks results, this appears to be more genuine.

Quote:

Originally Posted by Jowy Atreides

The 5870 in some games where ATI is favoured already bests the 480.
Since SI strengthens the weakest areas of 5 series performance and also includes a more efficient 4 shader model (and more of them) with a higher memory amount AND bandwidth (as well as more tesselation modules) then it's not very likely that the 6870 will be anything less than the 480 in the worst case scenario.

Definitely faster.

Who says the 4 D shader rumors are true those. This could simply be 1920 5 d shader like the previous gen which would yield between 10-20% more performance, but even that I think would be too conservative for AMD. However, all AMD needs to do right now to slap NV around is a conservative addition, as their current generation is beating NV pretty bad except for the gtx 460 when you include power consumption. If we look at it from that perspective, the performance falls exactly in line with this chart.

Their is one conflict however with the above chart, If AMD is going to have a 384 bit bus, why would they need ddr clocked at 1.5(6ghz effective) ghz? Thats too much bandwidth for any card.
08-31-2010, 09:39 PM
saaya

Quote:

Originally Posted by Jowy Atreides

The 5870 in some games where ATI is favoured already bests the 480.
Since SI strengthens the weakest areas of 5 series performance and also includes a more efficient 4 shader model (and more of them) with a higher memory amount AND bandwidth (as well as more tesselation modules) then it's not very likely that the 6870 will be anything less than the 480 in the worst case scenario.

Definitely faster.

well... on average a 480 IS faster than a 5870, and its at least 10%, last i checked it was around 15% iirc...

and we dont know how well ati tweaked the core and if it really gives a steady boost across the board... i agree, its likely that they wont settle for anything less than 480 performance... but im not sure if they will be fast enough to really beat the 480, as in enabling settings and resolutions the 480 cant...

Quote:

Originally Posted by tajoh111

For some reason, I think this is more about pricing than anything. It would seem too screwed up if the 6850 never caught up to the 5870 when there is only a 15 percent between the 5850 and the 5870. It would have to be like only 10% faster and to call that a 6850 would be shameful and misleading.

It would kind of mirror what NV has done with gtx 285 and gtx 275 those and could simply be a super minor update.

But only 10% difference. This could be fake and all but considering the timing of this lineup and the obvious flaws found in some of the benchmarks results, this appears to be more genuine.

well, despite the fake stuff chiphell has posted in the past i still trust them more than that weird chinese website that leaked those very high benchmark results... and hey, you know me, if i have to chose between optimistic and pessimistic rumors, which one do you think ill go for? ^^

jokes aside, i think the chiphell table makes a lot more sense than the numbers leaked by that other site... seems more realistic... im not saying i believe them... i still have my fingers crossed for bart reaching 5850 performance... :D
08-31-2010, 09:50 PM
tajoh111

Quote:

Originally Posted by saaya

well... on average a 480 IS faster than a 5870, and its at least 10%, last i checked it was around 15% iirc...

and we dont know how well ati tweaked the core and if it really gives a steady boost across the board... i agree, its likely that they wont settle for anything less than 480 performance... but im not sure if they will be fast enough to really beat the 480, as in enabling settings and resolutions the 480 cant...

well, despite the fake stuff chiphell has posted in the past i still trust them more than that weird chinese website that leaked those very high benchmark results... and hey, you know me, if i have to chose between optimistic and pessimistic rumors, which one do you think ill go for? ^^

jokes aside, i think the chiphell table makes a lot more sense than the numbers leaked by that other site... seems more realistic... im not saying i believe them... i still have my fingers crossed for bart reaching 5850 performance... :D

http://www.4gamer.net/games/110/G011...093/TN/002.jpg

The chiphell roadmap, matches this official roadmap more.

The things that matches in this case is the memory bus, both of which are 384 bits.

The thing that really sticks out as fake in those benches is the 256 bit bus in the GPU Z which doesn't correspond to anything mentioned officially so far.

With a 384 bit bus, all AMD needs is the DDR5 they are already using and not this super new 1.6ghz ddr5.
08-31-2010, 10:46 PM
kadozer

sorry have to post this from semiaccurate just take it as face value:

PHP Code:

New GPU family pricing structure: (pricing relative to the launch price of the Radeon HD 5000 series cards) Code: Antilles: Radeon HD 6970 ---------- N/A Radeon HD 6950 ---------- Radeon HD 5970 ($599) Cayman: Radeon HD 6870 ---------- N/A N/A ---------- Radeon HD 5870 ($399) Radeon HD 6850 ---------- N/A Radeon HD 6830 ---------- Radeon HD 5850 ($299) Barts: Radeon HD 6770 ---------- Radeon HD 5830 ($239) Radeon HD 6750 ---------- Radeon HD 5770 ($159) N/A ---------- Radeon HD 5750 ($109-129) Turks: Radeon HD 6670 ---------- Radeon HD 5670 ($99) N/A ---------- Radeon HD 5570 ($79-85) Radeon HD 6650 ---------- Radeon HD 5550 ($69) Caicos: Radeon HD 6xx0 ---------- Radeon HD 5450 ($49-59) Radeon HD 6xx0 ---------- N/A Memory bus width: * Antilles/Cayman 384-bit GDDR5 * Barts 256-bit GDDR5 * Turks 128-bit GDDR5 * Caicos 64-bit DDR3 All variants has UVD 3.0. Performance wise, Radeon HD 6770 will be slightly faster than GTX 460. Cayman PRO (Radeon HD 6850?): * PCB is slightly longer than Radeon HD 5870 (by a small margin) * [B]Card TDP <= 200 Watts[/B] * 850 MHz core frequency at this stage (may subject to change on the retail version) * comes equipped with a vapor-chamber cooler with uneven surface, looks like it would be expensive to make

So 6870 will be ~ 50w more than the 5870. (pure conjecture) hot hot hot

http://www.semiaccurate.com/forums/s...&postcount=333

Good news is we don't have to wait 6 months to find out.
08-31-2010, 10:51 PM
hurleybird

Quote:

Originally Posted by Zloyd

Is this reliable at all ? I mean 6850 to not perform as well as a 5870 ?

That just sounds wrong , well maybe its a little overeaction because it could perform almost as good at a signifiantly lower price and the same with 6870 matching 480 but lower prices.

It's obviously not comparing performance, but pricing. First, assuming the road map is legit, AMD would not show their hand by spilling performance like that. Secondly, no way a 6850 based on a part large enough to house a 384-bit memory bus is going to be slower than a 5870.

On one hand, I'm glad to see this alleged 384-bit bus version. Sounds like a monster, a cypress core enhanced and then scaled up closer to GF100 level proportions. In fact, this makes things look like the full design NI was just redone on 40nm despite the cost in die size (however, the fact that 6970 is supposed to come so soon afterwards seems to contraindicate this unless AMD has pulled a rabbit out of its hat in terms of energy efficiency or 6970 uses heavily cut down parts). On the other hand, we won't be able to see a 2GB version, only 1.5GB, and most definitely not 3GB.
08-31-2010, 11:35 PM
spursindonesia

@Saaya, i think that Chiphell roadmap is more related to pricing structure, agreed with Tajoh111 there. IMHO, this new generation will imply a higher price structure than the current Evergreen one (a trend that i loathe as a consumer but quite understandable from ATi POV since they want to maximize profit and the cost of goods must also increase in the process, chip is bigger & the specs seem more complex). There's still a hope for a 5850's capable Bart, because if you think Chiphell's leak is quite believable, then the buswidth design of each respective chips seems collaborative with the speculated relatively significant performance increase in this new family.
08-31-2010, 11:58 PM
RaZz!

the roadmap is definately based on prices, imo. neither ati nor amd ever used a roadmap like this to categorize performance.

since we saw no price drops in the 5870 era, i pretty much expected the 6870 to retail at a higher price point :/ especially if we consider they'll release the cards before christmas. maybe we'll see price drops after christmas already - atleast i hope so :p:

@saaya: your naming scheme doesn't make any sense tbh :p: ;) i'm confident the performance of these new babies will justify their placement in a 6000 series.
09-01-2010, 12:05 AM
Sn0wm@n

Quote:

Originally Posted by saaya

hmmm that doesnt sound that good...
more realistic, but a bit dissapointing...
looks like the 6870 will be close to a gtx480, but definitely not faster... might not even match it...

im really dissapointed that barts doesnt reach 5850 performance...
that would have been killer... a 150-200$ 5850...
it looks like the 6000 series will be a mere update, if this positioning is correct.
this looks like it doesnt deserve the 6000 name imo, according to this table it should rather be:

5990
5980

5890
5880
5860

5790
5780

how can a slide with names indicates that this wont beat fermi ????
09-01-2010, 12:19 AM
saaya

if ati thinks people will pay more money for faster cards, they dont seem to understand the market very well...
why do you think all dx11 cards combined only make up 5% of the market today?
cause games hardware requirements have stagnated ever since crysis came out... sure, there are one or two exceptions, but overall hardware requirements have not really increased over the past years...
you can play games on dx10 and some even on dx9 and you wont notice a difference...

more performance for higher prices is not going to be very successful i think...
what the market really wants is the same performance for lower prices... i thought ati had realized that...
the s/a rumors sound like the 6k series is going to boost performance, but there wont be any performance boosts for free, the price per fps seems to remain almost the same (bart xt = cypress le = gf104)
same perf, same price... meh...
09-01-2010, 12:23 AM
tajoh111

Quote:

Originally Posted by spursindonesia

@Saaya, i think that Chiphell roadmap is more related to pricing structure, agreed with Tajoh111 there. IMHO, this new generation will imply a higher price structure than the current Evergreen one (a trend that i loathe as a consumer but quite understandable from ATi POV since they want to maximize profit and the cost of goods must also increase in the process, chip is bigger & the specs seem more complex). There's still a hope for a 5850's capable Bart, because if you think Chiphell's leak is quite believable, then the buswidth design of each respective chips seems collaborative with the speculated relatively significant performance increase in this new family.

I think they need more ROPS those. Look how starved the 5830 is with 16. I think 32 is just enough for the 5870. I think its one of the reason the 5850 and 5870 don't have a bigger difference. With the 6870 they are likely going to add more.

I don't think the 256 bit bus is real though because it doesn't match the official roadmap and thus those benches are fakes, their is just too many flaws overall. Even the original poster implied that they were not genuine.

It doesn't mean the real thing won't live up to those results, it's just that some guy on the internet wanted to get attention and screwed with us a bit, and just posted a few bogus results.

I expect the 6870 to be fast, maybe a bit faster than a gtx 480, but I won't be optimistic. I remember thinking the 5870 was going to be faster than a 4870x2(basically a 4870x2 but with perfect scaling). It was atleast 25 percent slower than that.

Same with fermi, I was expecting a jump to gtx 285 SLI performance. Considering the increase in shaders and the higher clock speed, but as stuff started to leak like the decreased TMU and the reduced clock speed my expectation quickly sank.

The only time I was pleasantly surprised with performance was the 4870 and the 8800gtx. Both of which I can't see happening because 40nm at TSMC is a real pain in the ass to work with and the chips they have already right now are too big.

Quote:

Originally Posted by saaya

if ati thinks people will pay more money for faster cards, they dont seem to understand the market very well...
why do you think all dx11 cards combined only make up 5% of the market today?
cause games hardware requirements have stagnated ever since crysis came out... sure, there are one or two exceptions, but overall hardware requirements have not really increased over the past years...
you can play games on dx10 and some even on dx9 and you wont notice a difference...

more performance for higher prices is not going to be very successful i think...
what the market really wants is the same performance for lower prices... i thought ati had realized that...
the s/a rumors sound like the 6k series is going to boost performance, but there wont be any performance boosts for free, the price per fps seems to remain almost the same (bart xt = cypress le = gf104)
same perf, same price... meh...

I think people will pay more for more performance. The economy is starting to recover as tech company like AMD and Intel to show much stronger sales.

Most 58xx chips sold for well above retail. The retail price of the 5850 was 259 dollars. For the most part, the 5850 has sold for decently over 300 dollars because of crappy supply. None of that extra money really was received by AMD and went to retailers.

AMD would be more than happy than to have regular sales where cards were not sold out in hours and make extra money with a higher sale price that they themselves benefited.

Saaya, you think AMD can be really that happy from a 33 million dollar quarters net profit from their GPU department which they paid 5.4 billion dollars for? With their current margins and supply, that the best they can hope for considering how pathetic NV has been for the last 10 months. AMD like their CPU division want to increase their margins because thats the key to having big quarters.

Even those directx 11 cards only represent 5 or 6 percent of the pie. You have to remember, only one company has been really selling directx 11 cards and they were supply constrained. An NV part which most people were waiting for considering their current marketshare overall, has not really delivered and thus, those sale have remained unfulfilled. If NV released a good part(high performance, good power and heat) in october as originally planned, you would see atleast 2 or 3 times the current directx 11 cards.

Right now supply is limited and scalping of videocards show this. The logical thing to do is increase prices so demands doesn't outright outstrip supply and scalping stops. That way they sell pretty much as many videocards(you might not have as many sold out periods) and yet because of the higher margins you make way more money.
09-01-2010, 12:25 AM
flopper

Quote:

Originally Posted by saaya

if ati thinks people will pay more money for faster cards, they dont seem to understand the market very well...
...

That was called the Nvidia way we make people pay for cards.
They understood the market perfectly, if we have the best card (fastest) we charge for it.

The market isnt made of billions of saaya...individuals:rofl:
09-01-2010, 12:36 AM
SimBy

Quote:

Originally Posted by saaya

if ati thinks people will pay more money for faster cards, they dont seem to understand the market very well...
why do you think all dx11 cards combined only make up 5% of the market today?
cause games hardware requirements have stagnated ever since crysis came out... sure, there are one or two exceptions, but overall hardware requirements have not really increased over the past years...
you can play games on dx10 and some even on dx9 and you wont notice a difference...

more performance for higher prices is not going to be very successful i think...
what the market really wants is the same performance for lower prices... i thought ati had realized that...
the s/a rumors sound like the 6k series is going to boost performance, but there wont be any performance boosts for free, the price per fps seems to remain almost the same (bart xt = cypress le = gf104)
same perf, same price... meh...

Well it didn't stop nV fanboys buying ultras for $800 ;) Theres always market for the latest and greatest, no matter the price.

In any case, I wouldn't put too much into this leaks, they may all be fake.
09-01-2010, 12:46 AM
ilkkahy

Companies function is to maximase profits and they have bunch of smart people working there who have really hardass-degrees about doing that. Especially this is the case with amd which has bottomless pit of depts. I think companies will sell products with "socialist pricing" only when they need to keep or grow market share. So its not very smart to make card for the masses if it prevents the customers to purchase card for the next 3 years. Customers must be kept little unsatisfied all the time unless they have put really a ton of money to their product (in which case the marketing has done its function). I think that expensive high-end products are always welcome as long as there are good options for the average joe too.
09-01-2010, 06:33 AM
herderien

theory crafting can produce 12 pages of comments on names found in a .ini file ^^

good to know :p
09-01-2010, 06:51 AM
Jowy Atreides

Quote:

Originally Posted by saaya

if ati thinks people will pay more money for faster cards, they dont seem to understand the market very well...
why do you think all dx11 cards combined only make up 5% of the market today?

They will, they will, they will and ..they will.

Fools on the forums I frequent have tri sli 480's to play wow and emulators.
Epeen is all about the ego and we know the number one reason people buy overpriced products, emotional reaction.

Also 5% of the market is PLENTY to make bank profits.
Check the sales figures for all 5 series cards over the last year.

Millions sold to OEM's. Remember the mass majority of comuter buyers couldn't even point out a cpu socket on a motherboard.
They buy whatever the OEM tells them to buy and OEM's like product updates as excuses to force repeat sales.

Show 100 post(s) from this thread on one page

All times are GMT -8. The time now is 07:21 PM.

XtremeSystems