AMD's Radeon HD 5870: Bringing About the Next Generation Of GPUs
by Ryan Smith on September 23, 2009 9:00 AM EST- Posted in
- GPUs
AA Image Quality & Performance
With HL2 unsuitable for use in assessing image quality, we will be using Crysis: Warhead for the task. Warhead has a great deal of foliage in parts of the game which creates an immense amount of aliasing, and along with the geometry of local objects forms a good test for anti-aliasing quality. Look in particular at the leaves both to the left and through the windshield, along with aliasing along the frame, windows, and mirror of the vehicle. We’d also like to note that since AMD’s SSAA modes do not work in DX10, this is done in DX9 mode instead.
|
AMD Radeon HD 5870
|
AMD Radeon HD 4870
|
NVIDIA GTX 280
|
| No AA | ||
| 2X MSAA | ||
| 4X MSAA | ||
| 8X MSAA | ||
| 2X MSAA +AAA | 2X MSAA +AAA | 2X MSAA + SSTr |
| 4X MSAA +AAA | 4X MSAA +AAA | 4X MSAA + SSTr |
| 8X MSAA +AAA | 8X MSAA +AAA | 8X MSAA + SSTr |
| 2X SSAA | ||
| 4X SSAA | ||
| 8X SSAA |
From an image quality perspective, very little has changed for AMD compared to the 4890. With MSAA and AAA modes enabled the quality is virtually identical. And while things are not identical when flipping between vendors (for whatever reason the sky brightness differs), the resulting image quality is still basically the same.
For AMD, the downside to this IQ test is that SSAA fails to break away from MSAA + AAA. We’ve previously established that SSAA is a superior (albeit brute force) method of anti-aliasing, but we have been unable to find any scene in any game that succinctly proves it. Shader aliasing should be the biggest difference, but in practice we can’t find any such aliasing in a DX9 game that would be obvious. Nor is Crysis Warhead benefitting from the extra texture sampling here.
From our testing, we’re left with the impression that for a MSAA + AAA (or MSAA + SSTr for NVIDIA) is just as good as SSAA for all practical purposes. Much as with the anisotropic filtering situation we know through technological proof that there is better method, but it just isn’t making a noticeable difference here. If nothing else this is good from a performance standpoint, as MSAA + AAA is not nearly as hard on performance as outright SSAA is. Perhaps SSAA is better suited for older games, particularly those locked at lower resolutions?
For our performance data, we have two cases. We will first look at HL2 on only the 5870, which we ran before realizing the quality problem with Source-engine games. We believe that the performance data is still correct in spite of the visual bug, and while we’re not going to use it as our only data, we will use it as an example of AA performance in an older title.
As a testament to the rendering power of the 5870, even at 2560x1600 and 8x SSAA, we still get a just-playable framerate on HL2. To put things in perspective, with 8x SSAA the game is being rendered at approximately 32MP, well over the size of even the largest possible single-card Eyefinity display.
Our second, larger performance test is Crysis: Warhead. Here we are testing the game on DX9 mode again at a resolution of 1920x1200. Since this is a look at the impact of AA on various architectures, we will limit this test to the 5870, the GTX 280, and the Radeon HD 4890. Our interest here is in performance relative to no anti-aliasing, and whether different architectures lose the same amount of performance or not.
Starting with the 5870, moving from 0x AA to 4x MSAA only incurs a 20% drop in performance, while 8x MSAA increases that drop to 35%, or 80% of the 4x MSAA performance. Interestingly, in spite of the heavy foliage in the scene, Adaptive AA has virtually no performance hit over regular MSAA, coming in at virtually the same results. SSAA is of course the big loser here, quickly dropping to unplayable levels. As we discussed earlier, the quality of SSAA is no better than MSAA + AAA here.
Moving on, we have the 4890. While the overall performance is lower, interestingly enough the drop in performance from MSAA is not quite as much, at only 17% for 4x MSAA and 25% for 8x MSAA. This makes the performance of 8x MSAA relative to 4x MSAA 92%. Once again the performance hit from enabling AAA is miniscule, at roughly 1 FPS.
Finally we have the GTX 280. The drop in performance here is in line with that of the 5870; 20% for 4x MSAA, 36% for 8x MSAA, with 8x MSAA offering 80% of the performance. Even enabling supersample transparency AA only knocks off 1 FPS, just like AAA under the 5870.
What this leaves us with are very curious results. On a percentage basis the 5870 is no better than the GTX 280, which isn’t an irrational thing to see, but it does worse than the 4890. At this point we don’t have a good explanation for the difference; perhaps it’s a product of early drivers or the early BIOS? It’s something that we’ll need to investigate at a later date.
Wrapping things up, as we discussed earlier AMD has been pitching the idea of better 8x MSAA performance in the 5870 compared to the 4800 series due to the extra cache. Although from a practical perspective we’re not sold on the idea that 8x MSAA is a big enough improvement to justify any performance hit, we can put to rest the idea that the 5870 is any better at 8x MSAA than prior cards. At least in Crysis: Warhead, we’re not seeing it.
Facebook
Twitter
RSS
327 Comments
View All Comments
dieselcat18 - Saturday, October 3, 2009 - link
It truly amazes me that AnandTech allows a Troll like you to keep posting...but there is always one moron that comes to a forum like this and shows his a** to the world...So we all know it to be you...nice work not bringing anything resembling an intelligent discussion to the table..Oh and please don't tell me what it is that I bring to the conversation...my thoughts about this topic have nothing to do with my reply to you about your vulgar manner and lack of respect for anyone that has a difference of opinion.Oh and as for you paper launch...well sites like Newegg were sold out immediately because of the overwhelming demand for this card and I'll bet you anything there are cards available and in good supply at this very moment...Why don't you take a look and give us all another update.....I guess having that big "L" stamped on your forehead sums it up.....
SiliconDoc - Wednesday, September 23, 2009 - link
No, they didn't, because the 5870's just showed up last night, 4 of them, and just a bit ago the ONE of them actually became "available", the Powercolor brand.The other three 5870's are NOT AVAILABLE but are listed....
So "ATI paper launch" is the key idea here (for non red roosters).
1:43 PM CST, Wed. Sept. 23rd, 2009.
---
Yes, I watched them appear on the egg last night(I'm such a red fanboy I even love paper launches)... LOL
crimson117 - Wednesday, September 23, 2009 - link
Current cheapest GTX 295 at Newegg is $469.99.http://www.newegg.com/Product/Product.aspx?Item=N8...">http://www.newegg.com/Product/Product.a...p;cm_re=...
B3an - Wednesday, September 23, 2009 - link
Ryan, on your AA page, you have an example of the unofficial Nvidia SSAA where the tree branches have gone missing in HL2. And say because of this it's not suitable for general use.But for both the ATI pics, on either MSAA or SSAA, the tree branches are missing as well. Did you not notice this? because you do not comment on it.
Either way it looks like ATI AA is still worse, or there is a bug.
Ryan Smith - Wednesday, September 23, 2009 - link
We used the same save game, but not the same computer. These were separate issues we were chasing down at the same time, so they're not meant to be comparable. In this case I believe some of the shots were at 1600x1200, and others were at 1680x1050. The result of which is that the widescreen shots are effectively back a bit farther due to the use of the same FOV at all times in HL2.As you'll see in our Crysis shots, there's no difference. I can look in to this issue later however, if you'd like.
chizow - Wednesday, September 23, 2009 - link
Really enjoyed the discussion of the architecture, new features, DX11, Compute Shaders, the new AF algorithm and the reintroduction of SSAA an ATI parts.As for the card itself, its definitely impressive for a single-GPU but the muted enthusiasm in your conclusion seems justified. Its not the definite leader for single-card performance as the 295 is still consistently faster and the 5870 even fails to consistently outperform its own predecessor, the 4870X2.
Its scaling problems are really odd given its internals and overall specs seem to indicate its just RV790 CF on a single die, yet it scales worst than the previous generation in CF. I'd say you're probably onto something thinking AMD underestimated the 5870's bandwidth requirements.
Anyways, nice card and nice effort from AMD, even if its stay at the top is short-lived. AMD did a better job pricing this time around and will undoubtedly enjoy high sales volume with little competition in the coming months with Win 7's launch, up until Nvidia is able to counter with GT300.
chizow - Wednesday, September 23, 2009 - link
Holy....lolI didn't even realize til I read another comment that Ryan Smith wrote this and not Anand/Derek collaboration. That's a compliment btw, it read very Anand-esque the entire time! ;-) Really enjoyed it similar to some of your earlier efforts like the 3-part Vista memory investigation.
formulav8 - Wednesday, September 23, 2009 - link
I wouldn't be surprised if most of us already knew what was going to take place with performance and what-not. But its still a nice card whether I knew the specs before its official release or not. (And viewed many purposely leak benches). :)Jason
PJABBER - Wednesday, September 23, 2009 - link
Another fine review and nice to see it hit today. Your reviews are one reason I keep coming back to AT!Unfortunately, at MSRP the 5870 doesn't offer enough for me to move past the 4890 I am currently using, and bought for $130 during one of the sales streaks a month or so ago. Will re-evaluate when we actually start seeing price drops and/or DX11 games hit the shelves.
wicko - Wednesday, September 23, 2009 - link
It would have been nice to see 4890 in CF against 5870 in CF. 500$ spent vs 800$ spent :p