Nvidia Fermi versus radeon 5800 benchmarks out!

Page 6 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Apocalypse23

Golden Member
Jul 14, 2003
1,467
1
0
No, I've seen 5850s selling for $350, especially those that are most attractive to enthusiasts. I spent 4 days looking for an Asus for <$300, and I looked everywhere. The average prices I saw were between $330-$360 for those that were "in stock". Personally, I love the hunt, so I'll spend quite some time trying to find what I want for the price I'm willing to pay, it's part of the fun.

By blind luck I happened to stumble onto some for $294 at Amazon, they'd just received 10 at one of their warehouses. And when I say they'd just gotten in, I mean they weren't there one minute and they popped up the next. By the time I had completed the order, in the span of 3-4 minutes, they were all gone.

There's only 2 left, order yours right now:

Asus 5850

I picked mine up from their outlet here in Canada, they ship to US also.
 

MagickMan

Diamond Member
Aug 11, 2008
7,537
3
76
$18 for shipping? No bleeding way. That's probably why there are any left there. Amazon had free S&H, and if you have a free Prime membership (I do) it's even nicer. Mine should be here in ~6 hours.
 
Last edited:

taltamir

Lifer
Mar 21, 2004
13,576
6
76
gtx470 90-95&#37;
hd5870 100%
gtx480 110-115%

Not very exciting without relevant pricing information.

why only these three cards? here, lemme help you by copy pasting the entire list:

PCB lengths:
HD5870 280mm 28.0cm, 11.02 inches (not including backplate & enclosure)
GTX480 267mm 26.7cm, 10.51 inches
GTX470 243mm 24.3cm, 9.56 inches

General performance:
gtx285 75%
hd5850 80%
gtx470 90-95%
hd5870 100%
gtx480 110-115%
hd5970 140%

tough sell for nvidia, the AMD 5 series is faster. on the plus side, this SHOULD lower prices on AMD 5 series cards.
 

Keysplayr

Elite Member
Jan 16, 2003
21,209
50
91
Well only if people care about those things. Physx isn't anything special right now and we don't know how good their tessellation is yet when other things are going on other then only tessellation.

I wonder how long this "argument" will continue about "only tesselation".
Are you under the impression that the only thing rendered in the Unigine Heaven Bench, is tesselation? Tesselation is geometry setup isn't it? Are there no more polygons or textures rendered while tesselation is on? Or is it the same?
 
Last edited:

dug777

Lifer
Oct 13, 2004
24,778
4
0
I wonder how long this "argument" will continue about "only tesselation".
Are you under the impression that the only thing rendered in the Unigine Heaven Bench, is tesselation? Tesselation is geometry setup isn't it? Are there no more pixels or textures rendered while tesselation is on? Or is it the same?

Based on the heavy/medium/light tesselation benchmarks for at GTX 480 and 5970 in the techpowerup link (my caveat, and nothing more), clearly it isn't the same, otherwise the GTX 480 wouldn't get it's ass handed to it in every situation other than the heavy one, where it hands the ass of the 5970 to it...

If you see what I am saying, some pretty simple critical thinking suggests that if a GTX 480 can beat a 5970 (a card that it seems entirely unlikely to beat in any other benchmarks), but only in a heavy tesselation situation, then a conclusion that fermi is pretty special at 'only tesselation' is not unreasonable...

Similarly, from this:

Unigine Heaven: 1920x1200 16xaf benchmark
4xAA [fps]: 29/22/27 (GTX470/HD5850/HD5870)
8xAA [fps]: 20/19/23 (GTX470/HD5850/HD5870)

Nothing else in the techpowerup benchmark suggests the GTX 470 will generally beat the 5870 (hence the 90-95&#37; of one bandied about). Again, critical thinking suggests fermi's architecture stomps at 'only tesselation'...

My 2c ;)
 

dug777

Lifer
Oct 13, 2004
24,778
4
0
A further comment that the techpowerup WIC graph appears to be mislabelled, both from the colours and the positions of the text (the entire frame attached to the words 'Catalyst' is offset upwards in line with the text '5870' in green, and the lines match the naming colours), it looks like they have inverted the 5970 and GTX 470 in the text commentary...
 

Keysplayr

Elite Member
Jan 16, 2003
21,209
50
91
Based on the heavy/medium/light tesselation benchmarks for at GTX 480 and 5970 in the techpowerup link (my caveat, and nothing more), clearly it isn't the same, otherwise the GTX 480 wouldn't get it's ass handed to it in every situation other than the heavy one, where it hands the ass of the 5970 to it...

If you see what I am saying, some pretty simple critical thinking suggests that if a GTX 480 can beat a 5970 (a card that it seems entirely unlikely to beat in any other benchmarks), but only in a heavy tesselation situation, then a conclusion that fermi is pretty special at 'only tesselation' is not unreasonable...

Similarly, from this:

Unigine Heaven: 1920x1200 16xaf benchmark
4xAA [fps]: 29/22/27 (GTX470/HD5850/HD5870)
8xAA [fps]: 20/19/23 (GTX470/HD5850/HD5870)

Nothing else in the techpowerup benchmark suggests the GTX 470 will generally beat the 5870 (hence the 90-95&#37; of one bandied about). Again, critical thinking suggests fermi's architecture stomps at 'only tesselation'...

My 2c ;)

And what does tesselation do?

Here are some quotes from HardOCP's Unigine Heaven Benchmark review:

http://www.hardocp.com/article/2009/11/06/unigine_heaven_benchmark_dx11_tessellation

"Enabling DX11 compared to DX10 was positive for performance. You can clearly see that DX11 is slightly faster than DX10 all things being equal. This is good news and gives us hope that DX11 won&#8217;t suffer the huge drain on performance that DX10 did at the beginning of its short lifespan. There was a noticeable drain on performance enabling Tessellation however; the extra polygons do eat up performance. The Benchmark was still above 30 FPS at this setting at least on the Radeon HD 5850 and Radeon HD 5870."

"These images are self explanatory; Tessellation has the ability to add a lot of detail. What you see is that in the images without Tessellation there are fewer polygons, they are very simple and flat. When Tessellation is enabled the engine is dynamically generating in real-time more polygons as you move through the environment. This effect cannot be seen in these still screenshots, but if you run this program with the wireframe enabled and watch it in motion you will see the geometry being created on the fly.

Tessellation is quite amazing, and it really does add detail. The dragon alone has so many polygons with Tessellation that it looks like a texture wrapping around it. I was also impressed how much better things like the rope look with Tessellation. The rope actually looks like a rope with Tessellation. This is just an insane amount of geometry, no wonder why there is a large performance drop. Obviously a developer might not want to use this much in-depth Tessellation since it causes a large performance drop."

Tesselation adds polygons. A lot of them. And if say for example an HD5870 and GTX470 are roughly equal without tesselation, then enabling it tanks the HD5870 far more than the GTX470, what does this tell you? It tells me that the huge polygon increase is handled better on the GTX470. This article also mentions that there are several games on the way utilizing this engine. Under NDA, they could not mention the titles.

Here is a snip from the Summary of the article:

"While this Heaven Benchmark application isn&#8217;t an actual game, it does give us a glimpse of how Tessellation might be used in games. If used in this manner the benefits can be quite noticeable and greatly improve the gaming experience."

So, this actual engine will be used in games. And it's up to the developer as to actually how much tesselation is used. The more they use, I think you might know which cards may struggle more. You can say that the Unigine Heaven Bench is an extreme case of tesselation usage, but it gives an indication that when the geometry in cranked up and polygons severely increase, GTX4xx series has a leg up.

This is a feature. Much like any other. For example, in the past gens, AMD cards suffer less from higher levels of AA. Just the way it is.

Imagine an in-game setting that has a slider to control the level of tesselation, or just turn it on or off? Depends on dev I suppose.

Thoughts Dug? ZeroCool?
 
Last edited:

at80eighty

Senior member
Jun 28, 2004
458
3
81
This is a feature. Much like any other.

That is my view too - If tessellation horsepower matters to someone, they can pick the 470.

As an aside - does anyone here know how many games this year would be implementing lots of tessellation? (subjective as that parameter is)
 

Kuzi

Senior member
Sep 16, 2007
572
0
0

For six months now Nvidia guys were saying that DirectX11 was not important. Why the sudden infatuation with DX11/Tessellation? Oh wait, Nvidia is about to release cards that support DX11, awesome.

I agree with you about tessellation that by increasing polygons per object, the detail improves so more stress is put on the GPU, but we can't judge from just one benchmark. We need more real game tests to really tell which is better.

The GTX480, with 50% more transistors than Cypress, 50% more memory, and 50% higher bandwidth, "should" end up faster, obviously. The issue now is how much faster? And how much it will cost?
 

dug777

Lifer
Oct 13, 2004
24,778
4
0
And what does tesselation do?

Here are some quotes from HardOCP's Unigine Heaven Benchmark review:

http://www.hardocp.com/article/2009/11/06/unigine_heaven_benchmark_dx11_tessellation

"Enabling DX11 compared to DX10 was positive for performance. You can clearly see that DX11 is slightly faster than DX10 all things being equal. This is good news and gives us hope that DX11 won&#8217;t suffer the huge drain on performance that DX10 did at the beginning of its short lifespan. There was a noticeable drain on performance enabling Tessellation however; the extra polygons do eat up performance. The Benchmark was still above 30 FPS at this setting at least on the Radeon HD 5850 and Radeon HD 5870."

"These images are self explanatory; Tessellation has the ability to add a lot of detail. What you see is that in the images without Tessellation there are fewer polygons, they are very simple and flat. When Tessellation is enabled the engine is dynamically generating in real-time more polygons as you move through the environment. This effect cannot be seen in these still screenshots, but if you run this program with the wireframe enabled and watch it in motion you will see the geometry being created on the fly.

Tessellation is quite amazing, and it really does add detail. The dragon alone has so many polygons with Tessellation that it looks like a texture wrapping around it. I was also impressed how much better things like the rope look with Tessellation. The rope actually looks like a rope with Tessellation. This is just an insane amount of geometry, no wonder why there is a large performance drop. Obviously a developer might not want to use this much in-depth Tessellation since it causes a large performance drop."

Tesselation adds polygons. A lot of them. And if say for example an HD5870 and GTX470 are roughly equal without tesselation, then enabling it tanks the HD5870 far more than the GTX470, what does this tell you? It tells me that the huge polygon increase is handled better on the GTX470. This article also mentions that there are several games on the way utilizing this engine. Under NDA, they could not mention the titles.

Here is a snip from the Summary of the article:

"While this Heaven Benchmark application isn&#8217;t an actual game, it does give us a glimpse of how Tessellation might be used in games. If used in this manner the benefits can be quite noticeable and greatly improve the gaming experience."

So, this actual engine will be used in games. And it's up to the developer as to actually how much tesselation is used. The more they use, I think you might know which cards may struggle more. You can say that the Unigine Heaven Bench is an extreme case of tesselation usage, but it gives an indication that when the geometry in cranked up and polygons severely increase, GTX4xx series has a leg up.

Thoughts Dug? ZeroCool?

Firstly I am not at all trying to be antagonistic :)

I appreciate the angle you are taking, and it's not technically incorrect by any means, under what you concede to be an 'extreme case of tesselation usage' the fermi cards are very impressive.

However, it doesn't take away from the clear overall tale those benchmarks tell, that those 'tesselation examples' don't meaningfully relate to the real world right now, as we can see (from these admittedly leaked benchmarks) that neither the GTX 480 or 470 performance under heavy tesselation is mirrored more generally against the 5870 as a baseline across anything like a suite of games.

So clearly, the unigine tesselation is allowing the fermi cards to unleash some particular (and very impressive) capacity to handle tesselation (and you say more generally to handle polygons, but at this stage we will have to take your word for that extrapolation outside of polygons applied via tesselation will allow fermi to demonstrate a similarly commanding benchmark advantage over the equivalent AMD cards).

To put it more bluntly, their real world performance in current games doesn't correlate remotely meaningfully with their very impressive performance in a benchmark that is designed to show off tesselation. At this stage it would seema bit pre-emptive to read too much into that one benchmark without seeing performance in actual games that utlise tesselation, would it not?

No more, no less :)

For what it is worth, from a personal perspective, I have over the years tended to the view that I prefer to buy cards that provide me with fps/$ value in current games. Future proofing can often be somewhat of a nonsense because by the time tesselation/sm3.0/dx9 is meaningfully mainstream, those cards are very likely to lack the necessary balls to deliver the experience you could get from a cheaper overall upgrade path...so it all hinges on how Nvidia prices these cards whether I go for a 5850 or a GTX 470 :)

EDIT: Interesting that on Dirt 2, i notice that the GTX 470 loses to the 5870 in DX11 mode. Again I remember some sneering things being said about Dirt 2's DX11 capability, but it's still a real world DX11 example.
 
Last edited:

Lonyo

Lifer
Aug 10, 2002
21,939
6
81
And what does tesselation do?

Here are some quotes from HardOCP's Unigine Heaven Benchmark review:

http://www.hardocp.com/article/2009/11/06/unigine_heaven_benchmark_dx11_tessellation

"Enabling DX11 compared to DX10 was positive for performance. You can clearly see that DX11 is slightly faster than DX10 all things being equal. This is good news and gives us hope that DX11 won’t suffer the huge drain on performance that DX10 did at the beginning of its short lifespan. There was a noticeable drain on performance enabling Tessellation however; the extra polygons do eat up performance. The Benchmark was still above 30 FPS at this setting at least on the Radeon HD 5850 and Radeon HD 5870."

"These images are self explanatory; Tessellation has the ability to add a lot of detail. What you see is that in the images without Tessellation there are fewer polygons, they are very simple and flat. When Tessellation is enabled the engine is dynamically generating in real-time more polygons as you move through the environment. This effect cannot be seen in these still screenshots, but if you run this program with the wireframe enabled and watch it in motion you will see the geometry being created on the fly.

Tessellation is quite amazing, and it really does add detail. The dragon alone has so many polygons with Tessellation that it looks like a texture wrapping around it. I was also impressed how much better things like the rope look with Tessellation. The rope actually looks like a rope with Tessellation. This is just an insane amount of geometry, no wonder why there is a large performance drop. Obviously a developer might not want to use this much in-depth Tessellation since it causes a large performance drop."

Tesselation adds polygons. A lot of them. And if say for example an HD5870 and GTX470 are roughly equal without tesselation, then enabling it tanks the HD5870 far more than the GTX470, what does this tell you? It tells me that the huge polygon increase is handled better on the GTX470. This article also mentions that there are several games on the way utilizing this engine. Under NDA, they could not mention the titles.

Here is a snip from the Summary of the article:

"While this Heaven Benchmark application isn’t an actual game, it does give us a glimpse of how Tessellation might be used in games. If used in this manner the benefits can be quite noticeable and greatly improve the gaming experience."

So, this actual engine will be used in games. And it's up to the developer as to actually how much tesselation is used. The more they use, I think you might know which cards may struggle more. You can say that the Unigine Heaven Bench is an extreme case of tesselation usage, but it gives an indication that when the geometry in cranked up and polygons severely increase, GTX4xx series has a leg up.

This is a feature. Much like any other. For example, in the past gens, AMD cards suffer less from higher levels of AA. Just the way it is.

Imagine an in-game setting that has a slider to control the level of tesselation, or just turn it on or off? Depends on dev I suppose.

Thoughts Dug? ZeroCool?

It tells you the huge polygon increase is handled better?
It doesn't tell you that the card is better at creating the huge polygon increase?

AMD's dedicated tessellator [on the 2900] is capable of tessellating up to 15x more data and it can work much faster and more efficiently than a geometry shader set to the same task.

Surely it would make more sense that ATI doesn't have as much hardware working on creating the extra detail which needs to be rendered, and given that the Heaven benchmark is a benchmark, it would seem sensible to assume the workload remains constant, so if NV can create the data faster, it would have greater performance, because ATI would not be less able to render the scene, but less able to create the scene in order to even try rendering it.

Since Ungine's benchmark is heavily tessellated, it would make sense that the tessellator is the limiting factor, although we won't really know for another couple of weeks.
It would also make sense from a general point of view that ATI cards would struggle if they are designed with 'realistic' levels of tessellation in mind (that is, they have less tessellation capability now because games now won't use it much) while Fermi is more forward thinking (given that it's the first NV tessellation card) so it has a much greater capacity for tessellation, while ATI is just re-using their older fixed function hardware which has been around for a few years, but should be 'sufficient' for early real world scenarios.

Of course everything up to this point is speculative, and waiting a couple of weeks should hopefully reveal the answer.
I'm sure most people would agree not to put too much worth into benchmarks which don't reflect real world scenarios, especially when we don't have control over how the results are being made, as it were.
 

Keysplayr

Elite Member
Jan 16, 2003
21,209
50
91
For six months now Nvidia guys were saying that DirectX11 was not important. Why the sudden infatuation with DX11/Tessellation? Oh wait, Nvidia is about to release cards that support DX11, awesome.

I agree with you about tessellation that by increasing polygons per object, the detail improves so more stress is put on the GPU, but we can't judge from just one benchmark. We need more real game tests to really tell which is better.

The GTX480, with 50% more transistors than Cypress, 50% more memory, and 50% higher bandwidth, "should" end up faster, obviously. The issue now is how much faster? And how much it will cost?

You can search back if you like, Kuzi, but I'm pretty sure I have never said was not important. Never said is was important either. And I'M not the one infatuated with tesselation. It seems the whole tech web IS though. Probably the reason there is SO much discussion about tesselation over the last few months.
Now that we've settled the personal preference mystery, I appreciate your comments in the 2nd and 3rd paragraphs.

2nd para: Did you mean games with or without tesselation? Both? I would agree of course.
But be cautious about this "one" benchmark. As the article states that I linked to, there are several games under NDA that are on the way that use Unigine. So this "one" benchmark, could actually mean something more that "just one benchmark". We'll have to wait and see as usual.

3rd para: All of these questions will be answered shortly. Not long now.
 

Keysplayr

Elite Member
Jan 16, 2003
21,209
50
91
It tells you the huge polygon increase is handled better?
It doesn't tell you that the card is better at creating the huge polygon increase?



Surely it would make more sense that ATI doesn't have as much hardware working on creating the extra detail which needs to be rendered, and given that the Heaven benchmark is a benchmark, it would seem sensible to assume the workload remains constant, so if NV can create the data faster, it would have greater performance, because ATI would not be less able to render the scene, but less able to create the scene in order to even try rendering it.

Since Ungine's benchmark is heavily tessellated, it would make sense that the tessellator is the limiting factor, although we won't really know for another couple of weeks.
It would also make sense from a general point of view that ATI cards would struggle if they are designed with 'realistic' levels of tessellation in mind (that is, they have less tessellation capability now because games now won't use it much) while Fermi is more forward thinking (given that it's the first NV tessellation card) so it has a much greater capacity for tessellation, while ATI is just re-using their older fixed function hardware which has been around for a few years, but should be 'sufficient' for early real world scenarios.

Of course everything up to this point is speculative, and waiting a couple of weeks should hopefully reveal the answer.
I'm sure most people would agree not to put too much worth into benchmarks which don't reflect real world scenarios, especially when we don't have control over how the results are being made, as it were.

No matter how it is sliced, there are still a boatload more polygons that have to be given attention to by the shader processors in both AMD and NV architectures. How it gets there "could" play a part I suppose, but as for using that in an argument, there isn't really enough data yet to make that a true or false statement. Yet. For both our arguments.
 

Grooveriding

Diamond Member
Dec 25, 2008
9,107
1,260
126
Going on current pricing, 480 will probably come in at $500 tops, if the 5870 maintains its $400 price tag, 470 will be $400. Can't see the 480 going for $600, at that point why not spend another $100 for a 5970 for 170% or so the performance of the 480.
 

RussianSensation

Elite Member
Sep 5, 2003
19,458
765
126
Going on current pricing, 480 will probably come in at $500 tops, if the 5870 maintains its $400 price tag, 470 will be $400. Can't see the 480 going for $600, at that point why not spend another $100 for a 5970 for 170% or so the performance of the 480.

This logic is sound, except 5970s are very hard to find.
 

Borealis7

Platinum Member
Oct 19, 2006
2,914
205
106
if anything, a pricey 4XX could give AMD the "green light" (pun intended ;) ) to price their card 10&#37; BELOW the 4XX price.

a bad fermi = consumer loses.
 

Lonyo

Lifer
Aug 10, 2002
21,939
6
81
No matter how it is sliced, there are still a boatload more polygons that have to be given attention to by the shader processors in both AMD and NV architectures. How it gets there "could" play a part I suppose, but as for using that in an argument, there isn't really enough data yet to make that a true or false statement. Yet. For both our arguments.

Well NV said they focused more on geometry improvements that shaders in Fermi, being that they increase geometry processing by 8x over G200 (which means almost a 4x 'real' increase given that they doubled functional units anyway).

To put things in perspective, between NV30 (GeForce FX 5800) and GT200 (GeForce GTX 280), the geometry performance of NVIDIA’s hardware only increases roughly 3x in performance. Meanwhile the shader performance of their cards increased by over 150x. Compared just to GT200, GF100 has 8x the geometry performance of GT200, and NVIDIA tells us this is something they have measured in their labs.

Given that, it seems like an ATI running a single fixed function piece of tech which has been around for a long time in the graphics world (i.e. multiple years) vs an NV focused on increasing geometry and (seemingly) touting a much improved tessellation performance over ATI, it would make sense that it would be at least a significant part of performance inhibition, if not the sole cause of ATI's performance deficit, rather than being a sole throughput/shader/polygon issue.

Plus it would make sense given NV's arguably more forward thinking architecture vs ATI's evolution of their previous one, where they more added in the extras needed for DX11 rather than designing for it.
 

McCartney

Senior member
Mar 8, 2007
388
0
76
Does anyone have experience with processor design and architecture?
I think evolucion8 does, but no one else. Until then please be quiet.

I am sick and tired of people who own AMD video cards coming in and bashing nVidia products due to some sort of marketing strategy or their tactics, get over it.

Now for the people who don't want to argue intangibles and actual hardware, I'm down to talk. I've read 6 pages of subjective nonsense and focus on features. I want to hear someone actually explain how horrible the shader architecture is from ATi. Every driver update optimizes their compiler for new games, the fact of the matter is that ATi video cards are so optimization-dependent it is ridiculous.

People were crying when their 3xxx cards were crap until proper drivers came out. I remember people hailing the 4xxx series as the OWNAGE of the GTX 280. Someone show me reviews where the 4870 makes the GTX 280 or 285 pound sand in a game that's been release within the past year? I'm looking at benchmarks for "premature" 5850/5870 drivers and they don't perform as awesome as you guys say they do.

I am waiting patiently to buy my video cards, but I will tell you one thing, I will not let people who have this agenda against companies affect my decision.

I will also tell you guys to show me how the ATI architecture is superior to the nVidia because it isn't. It's a big bloody lie, the ATi architecture has 1 TRUE shader per cluster, and this shader is responsible for offloading the computation to the other 4 that are connected in series. You realize this is no different than the processor in our computers handling interrupts and sending instructions to the respective parts of the machine, right?

With that comes overhead, if you don't optimize the decoding of the instructions properly, you're asking for trouble.

Anyways, anyone want to argue the real meat here? Stop cowering and using horrible metrics to determine the superior card. Lets talk about the compiler, the hardware, the architecture; the guts!
 

Stoneburner

Diamond Member
May 29, 2003
3,491
0
76
Does anyone have experience with processor design and architecture?
I think evolucion8 does, but no one else. Until then please be quiet.

I am sick and tired of people who own AMD video cards coming in and bashing nVidia products due to some sort of marketing strategy or their tactics, get over it.

Now for the people who don't want to argue intangibles and actual hardware, I'm down to talk. I've read 6 pages of subjective nonsense and focus on features. I want to hear someone actually explain how horrible the shader architecture is from ATi. Every driver update optimizes their compiler for new games, the fact of the matter is that ATi video cards are so optimization-dependent it is ridiculous.

People were crying when their 3xxx cards were crap until proper drivers came out. I remember people hailing the 4xxx series as the OWNAGE of the GTX 280. Someone show me reviews where the 4870 makes the GTX 280 or 285 pound sand in a game that's been release within the past year? I'm looking at benchmarks for "premature" 5850/5870 drivers and they don't perform as awesome as you guys say they do.

I am waiting patiently to buy my video cards, but I will tell you one thing, I will not let people who have this agenda against companies affect my decision.

I will also tell you guys to show me how the ATI architecture is superior to the nVidia because it isn't. It's a big bloody lie, the ATi architecture has 1 TRUE shader per cluster, and this shader is responsible for offloading the computation to the other 4 that are connected in series. You realize this is no different than the processor in our computers handling interrupts and sending instructions to the respective parts of the machine, right?

With that comes overhead, if you don't optimize the decoding of the instructions properly, you're asking for trouble.

Anyways, anyone want to argue the real meat here? Stop cowering and using horrible metrics to determine the superior card. Lets talk about the compiler, the hardware, the architecture; the guts!

You have a general outline of a point but.... ummm why not measure with benchmarkery?
 

at80eighty

Senior member
Jun 28, 2004
458
3
81
that post reminds me of some kid who gets bullied all through HS, then listens to a Switchfoot song, gets inspired to change his life - then proceeds to go on a school shooting spree
 

SlowSpyder

Lifer
Jan 12, 2005
17,305
1,001
126

I don't have experience with processor design, I guess I can't post according to you, but if I have your permission I'll make this post... since you're making the rules on who can post in this thread.

If any Nvidia bashing is going on it's because right now from the enthusiast gamer point of view they are getting pummeled (This bashing also occured for AMD/ATI back in the Radeon 2900XT days). Regardless of who's architecture is inferior/superior AMD has a complete line of DX11 parts out and are better at almost every price point as well as having the fastest parts with some features that some gamers do want that can only be had on AMD hardware (much like Physx and 3D glasses can only be had on Nvidia hardware).

I don't recall anyone here saying a 4870 would be faster than a GTX280, though I do recall people claiming it may even match it. I also remember OCGuy smuggly putting that as his sig, when someone claimed the 4870 may even match the GTX280. His sig changed shortly after launch. I think most people were excited about getting 80-90&#37; of the performance of a $650 GTX280 for $299. Again, this is regardless of who has what architecture. If AMD's architecture was so terrible I doubt Nvidia would have cut the GTX280 price in half.

You can argue one architecture is better than another depending on the metric you use. It's not worth arguing about as it's been beaten to death in the past and depending on how you look at it one can be better than the other.

Well, I hope you don't mind if myself or other people continue to post even if we don't have processor design experience. Good luck with your superior card that doens't yet exist except on paper. The superior card that will likely be quite hard to find for quite a while after launch because it's superior design appears to be very difficult to make yield well at this point (according to a few rumors floating around). I don't think most of us have a fanboy issue, most of us talk with our wallets. What would you rather game on, a $355 GeForce GTX285 or a $310 Radeon 5850?
 
Last edited:

Kuzi

Senior member
Sep 16, 2007
572
0
0
You can search back if you like, Kuzi, but I'm pretty sure I have never said was not important. Never said is was important either. And I'M not the one infatuated with tesselation. It seems the whole tech web IS though. Probably the reason there is SO much discussion about tesselation over the last few months.
Now that we've settled the personal preference mystery, I appreciate your comments in the 2nd and 3rd paragraphs.

You didn't say it Keys, but I mean many others did. It will be a while before developers really start pushing DirectX11 features, and next-gen cards will be out by that time. That's what many (mostly NV fans) were saying for the past six months, and it is true to some extent. Only a few DX11 games are out now, but I'm sure many more will come out this year.

2nd para: Did you mean games with or without tesselation? Both? I would agree of course.
But be cautious about this "one" benchmark. As the article states that I linked to, there are several games under NDA that are on the way that use Unigine. So this "one" benchmark, could actually mean something more that "just one benchmark". We'll have to wait and see as usual.

From what I understand, tessellation will be scalable. So a game can run with tessellation on lower end cards with less complex models, or on high end cards with more complexity. It all happens on the fly, no need to change any settings. I think the Unigine benchmark has the tessellation complexity predefined/static for benchmarking purposes, but games using tessellation will increase/decrease world complexity dynamically.This may make it harder to notice the difference between NV and ATI hardware tessellation in actual games, unless a screen shot was taken. I could be wrong about this though.

Nvidia may have the better tessellation implementation right now, Anand himself talked about that, so lets wait and see. And hopefully these Unigine games show some great tessellation effects, although by the time they release Northern Islands may be out, which may offer an improved tessellator.
 

Lonyo

Lifer
Aug 10, 2002
21,939
6
81
I am sick and tired of people who own AMD video cards coming in and bashing nVidia products due to some sort of marketing strategy or their tactics, get over it.

People were crying when their 3xxx cards were crap until proper drivers came out. I remember people hailing the 4xxx series as the OWNAGE of the GTX 280. Someone show me reviews where the 4870 makes the GTX 280 or 285 pound sand in a game that's been release within the past year? I'm looking at benchmarks for "premature" 5850/5870 drivers and they don't perform as awesome as you guys say they do.

I am waiting patiently to buy my video cards, but I will tell you one thing, I will not let people who have this agenda against companies affect my decision.

I will also tell you guys to show me how the ATI architecture is superior to the nVidia because it isn't. It's a big bloody lie, the ATi architecture has 1 TRUE shader per cluster, and this shader is responsible for offloading the computation to the other 4 that are connected in series. You realize this is no different than the processor in our computers handling interrupts and sending instructions to the respective parts of the machine, right?

With that comes overhead, if you don't optimize the decoding of the instructions properly, you're asking for trouble.

Anyways, anyone want to argue the real meat here? Stop cowering and using horrible metrics to determine the superior card. Lets talk about the compiler, the hardware, the architecture; the guts!

How is the ATI architecture superior? One simple way: It does less, with less.
I don't think anyone has ever claimed the HD4870 beats the GTX280 or GTX285, but that doesn't mean it doesn't "own" it. It's quite significantly cheaper and performance is pretty close.
When it comes to architecture, technical minutia don't really matter, what matters is how it performs, and the HD4800 cards manage 90+% of the GTX27/8 cards typically using about 70% of the transistors.
I'd say that's kind of superior. More work per transistor. Sure, they have less of them, but that's more design philosophy than anything.
 

GaiaHunter

Diamond Member
Jul 13, 2008
3,628
158
106
I wonder how long this "argument" will continue about "only tesselation".
Are you under the impression that the only thing rendered in the Unigine Heaven Bench, is tesselation? Tesselation is geometry setup isn't it? Are there no more polygons or textures rendered while tesselation is on? Or is it the same?

But then the GTX 470 and 480 would be much faster than the 5870 in almost situations. Assuming all these leaked benchmarks are correct, that doesn't seem the case.