ATI 4xxx Series Thread

[firewolfsm]

Then we can tune down the physics effects just a little and run it on CPUs. I'd rather have a high resolution.

 

[biostud]

Originally posted by: Continuity28

Originally posted by: Extelleron
How do you plan on diverting shading resources needed for graphics processing to physics processing without reducing performance?

And I'm sure code optimized for an Octal-core processor would fare well against a modern video card.

The point is, running such physics calculations on the underloaded CPU would result in less FPS than running it on the GPU, even when the GPU is already fully loaded with tasks. It's because the CPU is very slow at this type of task. When the GPU does it, you may drop in FPS from 65 to 62, or hell even for argument, 45. When the CPU does it, you may drop from 65 to 10 because the CPU can't do that level of physics any faster than that.

You can't really optimize the code for the CPU, the issue starts at the hardware level, the hardware isn't built for it. There are things that the modern GPU can do better than a CPU, and things a modern CPU can do better than a GPU, and that's why they have their respective uses and roles in a system. We don't run our operating systems and programs on a GPU, we don't render modern games on our CPU and expect speed. They are drastically different pieces of hardware.

It's like comparing an elephant and horse as modes of transportation. The elephant can carry more for sure, but it's not as fast regardless of it's capacity. If you think the horse will always be faster, try loading it as much as the elephant and see how fast it runs. They aren't built the same way, and won't be used in the same ways.

I don't care if anyone hates my analogy.

It depends on the amount of physics calculations needed in a game. I seriously doubt that we will see games needing a PPU for a long time. I think that using a multicore CPU for physics is the way to go for game developers, because "everyone" has one. Not that many has a SLI/CF setup.

Of course you can run a tech demo that cripples the CPU, in physics, but if you want to get the best looking techdemo you would more likely use >90% of the GPU for graphics. The same goes for games, better graphics sells better than an advanced physics system.

But when we get an integrated GPU/CPU then it would make much sense to run the physics on the GPU, or before that run it on the IGP.

 

[Denithor]

Originally posted by: Continuity28
You can't really optimize the code for the CPU, the issue starts at the hardware level, the hardware isn't built for it. There are things that the modern GPU can do better than a CPU, and things a modern CPU can do better than a GPU, and that's why they have their respective uses and roles in a system. We don't run our operating systems and programs on a GPU, we don't render modern games on our CPU and expect speed. They are drastically different pieces of hardware.

So how/will Nehalem with GPU integrated into the CPU core affect this issue? Perhaps the code could be optimized so that core handles physics calcs efficiently while the other CPU cores handle driving the system and the discrete GPU handles graphics?

 

[Martimus]

Originally posted by: biostud
The thing is most people prefer good loking graphics compared to intense physics calculation.

I actually prefer better physics to AA or HDR or such graphical options. Proper physics really adds to the realism of the game more than any other effect I have experienced.

 

[taltamir]

its funny that you say that physx aren't graphics, so far the ONLY physx I saw was graphics... realistic smoke and clothe are pure graphics and have nothing to do with gameplay, destructible environment and lethal shrapnel are possible, but have yet to be utilized.
Although there is the "tornado level"... i guess that one would count as game effects.

it is really a matter of "how much"... if adding physx drops you from 65fps to 62fps on GPU or 10fps on CPU (realistic with what I know on the respective power of the hardware) then it makes a lot of sense.
If adding it drops you from 65fps to 20fps on GPU and 30fps on CPU then it makes no sense...

One thing though, can the physx engine potentially render physx on both the GPU and CPU? That would mean that it WILL max out the CPU, and then use the GPU for a little boost, and it will always be welcome.

PhysX type graphic enhancements are the way of the future. Honestly crysis didn't impress me all that much, insane GPU demand and it gave so little... but realistic looking clothe and liquids, that tear and splatter? that is all eye candy physx (has no in game effect) and it looks BEAUTIFUL!

PS. I loved the horse and the elephant analogy, sure the horse is faster, but try trying it to a whole tree, the horse would not be able to move, while the elephant will uproot the tree and keep on marching at the same pace dragging it behind him. Probably the best analogy of CPU vs GPU.

 

[dv8silencer]

Yeah.. realism is not only how close, graphically, the game looks like reality, but also how it behaves!

 

[Janooo]

R700 lost battle to GTX 280?

We learn from pcinlife forum that ATI next generation flagship R700 lost battle to GTX 280 in closed doors tests.

Taiwan graphics cards manufacture obviously got both 4870x2 and GTX 280 graphics card samples for these comparisons. But as we told you before, we still don't know the specification of the test hardware and software configurations.

We hope ATI Catalyst Develop Team could do miracle to save Radeon HD 4870x2 in next gorgeous chips battle between AMD and NVIDIA.

 

[HOOfan 1]

all of these pre-release "benchmarks" are pretty much worthless...especially when it isn't first hand information. All of the "benchmark" reports start with a "we heard from so and so, who heard from so and so, who heard from a friend that works an an ATI board partner"

I remember when bittech posted some "benchmarks" for the 9800GX2 several days before it was release that showed it slower than the 3870X2 and then it came out and stomped the 3870X2.

I will wait until both cards are out and multiple sites are quoting their own results before I make a final judgement

 

[Quiksilver]

Originally posted by: Janooo
R700 lost battle to GTX 280?

We learn from pcinlife forum that ATI next generation flagship R700 lost battle to GTX 280 in closed doors tests.

Taiwan graphics cards manufacture obviously got both 4870x2 and GTX 280 graphics card samples for these comparisons. But as we told you before, we still don't know the specification of the test hardware and software configurations.

We hope ATI Catalyst Develop Team could do miracle to save Radeon HD 4870x2 in next gorgeous chips battle between AMD and NVIDIA.

That was a terrible article, no numbers, no anything factual. Just rumors, what a bore.

 

[Janooo]

Originally posted by: Quiksilver

Originally posted by: Janooo
R700 lost battle to GTX 280?

We learn from pcinlife forum that ATI next generation flagship R700 lost battle to GTX 280 in closed doors tests.

Taiwan graphics cards manufacture obviously got both 4870x2 and GTX 280 graphics card samples for these comparisons. But as we told you before, we still don't know the specification of the test hardware and software configurations.

We hope ATI Catalyst Develop Team could do miracle to save Radeon HD 4870x2 in next gorgeous chips battle between AMD and NVIDIA.

That was a terrible article, no numbers, no anything factual. Just rumors, what a bore.

Here you go. Some numbers

 

[Quiksilver]

Originally posted by: Janooo

Originally posted by: Quiksilver

Originally posted by: Janooo
R700 lost battle to GTX 280?

We learn from pcinlife forum that ATI next generation flagship R700 lost battle to GTX 280 in closed doors tests.

Taiwan graphics cards manufacture obviously got both 4870x2 and GTX 280 graphics card samples for these comparisons. But as we told you before, we still don't know the specification of the test hardware and software configurations.

We hope ATI Catalyst Develop Team could do miracle to save Radeon HD 4870x2 in next gorgeous chips battle between AMD and NVIDIA.

That was a terrible article, no numbers, no anything factual. Just rumors, what a bore.

Here you go. Some numbers

I know those numbers. The GT200 performing better than 4870X2 had none, didn't say by how much, no hardware, etc. It was all fud. Even Fudzilla puts some guesswork numbers.

 

[taltamir]

I agree, so far it is just wild speculations. We are going to have to wait some more for actual info.

 

[deizel]

When is the 4870x2 suppose to be out?

 

[Rusin]

Late Q3

 

[MarcVenice]

http://www.nordichardware.com/news,7755.html

Some new info, for a nice price this will prove to be pretty interesting if you ask me.

 

[ghost recon88]

Originally posted by: MarcVenice
http://www.nordichardware.com/news,7755.html

Some new info, for a nice price this will prove to be pretty interesting if you ask me.

Hmm most interesting. At only $250 a pop for a HD4870, it will be hard to beat.

 

[Continuity28]

Originally posted by: ghost recon88

Originally posted by: MarcVenice
http://www.nordichardware.com/news,7755.html

Some new info, for a nice price this will prove to be pretty interesting if you ask me.

Hmm most interesting. At only $250 a pop for a HD4870, it will be hard to beat.

It's looking very attractive.

I can't wait for benchmarks and user reports of compatibility with new and older games.

 

[ddarko]

Originally posted by: ghost recon88

Hmm most interesting. At only $250 a pop for a HD4870, it will be hard to beat.

I think you misread the chart. It's $250 for the 4850; the 4870 is $100 more.

In any case, I think with the launch so close and ATI leaking like a dutch dike, all the reports that the 4870 is slower than the GTX 280 is very plausible and right. It shouldn't come as a surprise since ATI long ago announced its intentions to use dual chip cards to address the high-end of the market. Why doesn't anyone believe it? It meant what it said, which is both good and bad, good because it's mainstream offerings are affordable, bad because those who don't like dual chip solutions have only one company that offers a high-end single GPU. Oh well.

edited by: ddarko

I just read the link to Nordic Hardware and see that it says the 4870 will be $250. Sorry about that. I got the Nordic link mixed up with another site that shows a purported chart from AMD that lists specs and the price of the 4870 as $350. I do think that a $250/350 price is much more plausible than the $200/250 prices suggested by Nordic (assuming all the sites are talking in American dollars).

 

[bryanW1995]

Originally posted by: AshPhoenix
RV770PRO engineering sample works at 625MHz

Product part to be faster

The RV770PRO is a GDDR3 version of the RV770 chip and we?ve found out that the samples out in the wild are working at 625MHz. The samples are based on revision a12, which is usually ATI?s production revision.

The sample board was equipped with GDDR3 memory clocked at 2,000MHz and we?ve heard that GDDR5 samples are expected any day now.

Both RV770XT and RV770PRO are scheduled for a June 16th launch, just two days before the Geforce GTX 280 / 260 launch.

fuad alert!!

 

[bryanW1995]

Originally posted by: Quiksilver

Originally posted by: Janooo

Originally posted by: Quiksilver

Originally posted by: Janooo
R700 lost battle to GTX 280?

We learn from pcinlife forum that ATI next generation flagship R700 lost battle to GTX 280 in closed doors tests.

Taiwan graphics cards manufacture obviously got both 4870x2 and GTX 280 graphics card samples for these comparisons. But as we told you before, we still don't know the specification of the test hardware and software configurations.

We hope ATI Catalyst Develop Team could do miracle to save Radeon HD 4870x2 in next gorgeous chips battle between AMD and NVIDIA.

That was a terrible article, no numbers, no anything factual. Just rumors, what a bore.

Here you go. Some numbers

I know those numbers. The GT200 performing better than 4870X2 had none, didn't say by how much, no hardware, etc. It was all fud. Even Fudzilla puts some guesswork numbers.

The gt200 SHOULD perform better than 4870x2 until the ati guys can code cat drivers for it. It's MUCH harder to write drivers for xfire. Wait until both cards are released and we'll have a better idea of the relative performance.

 

[taltamir]

bryan makes a good point. You would have to wait a few months for the 4870x2 CF drivers to mature before it starts competing with (or even defeating) the G200... And you will have to wait again for every new release.

 

[dv8silencer]

http://www.dailytech.com/article.aspx?newsid=11842

I thought the 4800s were going to be either GDDR3 or GDDR5 ?

 

[Elfear]

Originally posted by: taltamir
bryan makes a good point. You would have to wait a few months for the 4870x2 CF drivers to mature before it starts competing with (or even defeating) the G200... And you will have to wait again for every new release.

According to rumor, I thought the R700 was supposed to implement a hardware-based CF solution so drivers won't be nearly as important as they are now.

 

[taltamir]

the rumer is that it supposedly IMPROVES the hardware implementation... not full hardware implementation, but a step beyond the current... isnt the current model something like 5th gen CF already?

Every generation slightly improves the efficiency and capability of multi GPU setups. but it still has a way to go.

 

[Rusin]

Bryan:
That 2nd Gen unified shaders on GT200 should be intresting; Nvidia claims 50% bigger performance with GT200 shader than with G80/g92 shader...