• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

AMD V Nvidia by Richard Huddy

Page 10 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Nvidia has a much more powerful solution for tesselation today. It would had been nice if AMD countered them so developers arent making games towards the lowest common denominator. The hardware always preceeds the software. So I dont buy into the argument about it not mattering now. Without a good hardware base now, 18-24 months from now games wont be as nice as they could be. Which hardware would get us to a point in the future faster? Nvidia or AMD?
It doesn't matter who gets us there faster as long as we get there.The same thing with hardware physics .....both of them have to support these things properly for them to be implemented in games.

However I don't think the Tessellation issue is enough to rain on AMD'S parade right now.These cards look to be excellent from what I've seen.
 
The same thing with hardware physics .....both of them have to support these things properly for them to be implemented in games.

By the looks of it, nVidia is doing fine getting hardware physics implemented in games, without AMD's support. You may not like the results, but developers ARE adding hardware physics to games, while there's only one IHV that supports it.

However I don't think the Tessellation issue is enough to rain on AMD'S parade right now.These cards look to be excellent from what I've seen.

The value is great, yes... It's just sad that AMD has not addressed this particular weakness in the 5000-series with this refresh. I also don't like the way AMD handles the situation.
 
Patience is what we need.Both companies will surely have better Tessellation in the future.But we have to ask if it really matters now.The prices on these cards seem okay to me.But as a 5850 owner I'll wait on a bigger upgrade.

It's always the same, when one IHV has a feature advantage -- it will hype it over the other and when it doesn't, will try to cast shadows on it and shine the light on what they do well.

PR guys, hehe, seem to get caught in 180's as their views change on a dime depending on their advantages and disadvantages; and change depending on what they're selling for a specific window of time. I don't get worked up over it and expect it -- it's their job.

A lot of extremists simply parrot but the rare ones are consistent over time.

I'm trying to find out why the developers are compelled to add the amplification for HawX?

Is it nVidia simply trying to find ways of offering differentiation with their tessellation over AMD's?

Does it really bring in improvements in IQ, while saving band with and smaller memory footprints -- that can be used for resolution and AA?
 
Then give AMD credit for finally having a GTX 460 competitor.The recent price shakedown by Nvidia is not a good thing??

I did, I said it's good value...
But this thread was more about Richard Huddy, and his attack on nVidia, tessellation etc.
 
Does it really bring in improvements in IQ, while saving band with and smaller memory footprints -- that can be used for resolution and AA?

It's such a shame that the Unigine Heaven benchmark can not be run with all the geometry statically pre-tessellated. If you turn tessellation off, you just get very low geometry detail.
If you could run the same level of detail with and without tessellation, you could compare the performance with equal IQ.
 
By the looks of it, nVidia is doing fine getting hardware physics implemented in games, without AMD's support. You may not like the results, but developers ARE adding hardware physics to games, while there's only one IHV that supports it.



The value is great, yes... It's just sad that AMD has not addressed this particular weakness in the 5000-series with this refresh. I also don't like the way AMD handles the situation.
Well give me a list of upcoming triple A titles that are supporting hardware Physics.So far I've not been impressed with any of the ten or so games in the last few years that had this feature.But that's a touchy topic that's been beaten to death anyway.

I think they will HAVE to address this Tessellation issue if they want to continue competing with Nvidia but for now it's essentially a non issue for today's and also near future games.
 
It doesn't matter who gets us there faster as long as we get there.The same thing with hardware physics .....both of them have to support these things properly for them to be implemented in games.

However I don't think the Tessellation issue is enough to rain on AMD'S parade right now.These cards look to be excellent from what I've seen.

Get where though? If we follow AMDs track we stop about a factor of 8-10. Nvidia much higher. Which is exactly my point.
 
But that's a touchy topic that's been beaten to death anyway.

Exactly.

I think they will HAVE to address this Tessellation issue if they want to continue competing with Nvidia but for now it's essentially a non issue for today's and also near future games.

I don't think it is.
I think nVidia has been working hard with all developers in the TWIMTBP program to put highly detailed tessellation in every game, since the release of Fermi.
As a result, I bet that quite a few games with heavy tessellation will be hitting the market soon.
I think Metro was already a taste of things to come (as I said, it wasn't done well, but regardless of that, it does put AMD at a disadvantage, and that's the main goal for nVidia), and HAWX 2 is another.
 
Get where though? If we follow AMDs track we stop about a factor of 8-10. Nvidia much higher. Which is exactly my point.
It's all a matter of balance.They can't lag behind in Tessellation for too long which is part of DirectX 11 BUT most people are looking for a balance between speed,IQ,cost,heat,power etc.When looking at the Tessellation thingy it might be a very small factor in their purchase decision in the big picture.
 
It's all a matter of balance.They can't lag behind in Tessellation for too long which is part of DirectX 11 BUT most people are looking for a balance between speed,IQ,cost,heat,power etc.When looking at the Tessellation thingy it might be a very small factor in their purchase decision in the big picture.

I think we already established this wont affect many purchasing the card today. What we are debating now is how AMD will hold back development of some games due to an inferior performance in a building block of DirectX 11 within their hardware.
 
Exactly.



I don't think it is.
I think nVidia has been working hard with all developers in the TWIMTBP program to put highly detailed tessellation in every game, since the release of Fermi.
As a result, I bet that quite a few games with heavy tessellation will be hitting the market soon.
I think Metro was already a taste of things to come (as I said, it wasn't done well, but regardless of that, it does put AMD at a disadvantage, and that's the main goal for nVidia), and HAWX 2 is another.
Can you name any others beside these two?
 
I think we already established this wont affect many purchasing the card today. What we are debating now is how AMD will hold back development of some games due to an inferior performance in a building block of DirectX 11 within their hardware.
We all know sometimes they take their own sweet time to do things but when the demand for hardware that can handle tessellation properly increases you bet they'll get cracking on it.

But I would like to know the upcoming games you are both speaking about.
 
How many developers do you know have ever had enough power from the hardware? That is a silly comment imo. And in this case considering the vastly superior performance of their competitor. I am sure they would rather work with that than AMDs solution.
Again, it's up to the developers to determine how much tessellation horsepower they really require. To say that Nvidia's tessellation power plenty while simultaneously saying that AMD doesn't provide enough seems rather shortsighted to me.

Each developer will have their own opinion about how much tessellation is "enough". Some will say neither AMD nor Nvidia is powerful enough to satisfy them. Some will say Nvidia has enough while AMD is lagging. Some will say both AMD and Nvidia tessellation levels are just fine for now. It's all just a matter of opinion that will vary from studio to studio.
 
Again, it's up to the developers to determine how much tessellation horsepower they really require. To say that Nvidia's tessellation power plenty while simultaneously saying that AMD doesn't provide enough seems rather shortsighted to me.

Each developer will have their own opinion about how much tessellation is "enough". Some will say neither AMD nor Nvidia is powerful enough to satisfy them. Some will say Nvidia has enough while AMD is lagging. Some will say both AMD and Nvidia tessellation levels are just fine for now. It's all just a matter of opinion that will vary from studio to studio.


Why is AMD then whining about that Ubisoft rejected their "optimized" tessellation code.
If it's up to the developer, AMD should let the topic go and accept the level of tessellation that Ubisoft think they need ...right?
 
Yup... that screenshot in there also shows some very heavy tessellation, it seems like they use far smaller triangles than the 16-pixels that AMD now wants (and that was on previous-gen hardware, not the current 6000-series with their super-duper improved tessellator).
Likewise AMD's own DX11 detail tessellation sample is suddenly the 'wrong' way to do tessellation... why? Just because nVidia does it a lot better than they do.

Just a small correction, that tessellation demo is actually done in DX9 and is rather old.
 
Now I uderstand why Fuddy cries like he does...that is a beatdown with a vengance...

Right, I would hate to be playing a benchmark of an unreleased game maxed out at 1920x1200 8xAA at 70FPS with unoptimized drivers (as AMD have said). The fact that a 5770 is within 20fps of a 5870 should show you that it needs some work.
 
unoptimized drivers (as AMD have said).

Yea right... It's standard DX11 code. AMD has had DX11 drivers out for about a year now, with plenty of updates. Lots of software with tessellation has been around, including various samples from AMD in the early days.
You think they STILL didn't get around to optmizing the tessellation? Nope, it's the hardware.

The fact that a 5770 is within 20fps of a 5870 should show you that it needs some work.

That's the same pattern we've been seeing in all heavy tessellation scenario's... The tessellator is bottlenecking the rest of the GPU. Same tessellator in the 5770 as in the 5870, hence same performance.
 
Back
Top