DiRT 2: DirectX 11 Game Performance Compared And Analyzed

[cleeve]

[citation][nom]p4l1ndr0m3[/nom]I mean, am I really missing something here? I thought the i920 systems raped anything AMD. Am I wrong?[/citation]

There are so many variables it's difficult to say without having the machines in the same lab. For example, when you say 'with everything on' are you implying AA as well? If so, how much AA? 2x? 4x? 8x? etc.

Other than that, the CPU benches show that the CPU isn't as much of a factor in this title. the i7 920 will show the biggest gains in CPU-limited titles, while DiRT 2 isn't CPU-limited at all. With this in mind, I'd expect similar graphical settigns on identical graphics cards to dfeliver identical performance, and that's basically what you've shown us here...

 

[hundredislandsboy]

[

I'm not at home, but I'll be happy to provide screenshots from the Dirt2 benchmark. I mean, am I really missing something here? I thought the i920 systems raped anything AMD. Am I wrong?[/quotemsg]


Uh yeah there was a rape and continued raping and Intel is doing it except they're not doing to AMD. They're doing it to Intel fanboys.

 

[rambo117]

damn.. look at those water effects

I'm more looking forward to Stalker Call of Pripyat; Tessellation looks incredible in that title.

 

[t1daballer]

This article is very misleading. First and foremost, the 5770 perform much higher than their showing in this article. Secondly, they used an outdated driver, 9.12 hotfix is the main one out and there's a big image difference between dx 11 and dx 9. Hardocp article is much more in depth, and is much better detailed than this shenanigan of an article. I have a core 2 quad Q9550 @ 4 ghz, 8Gb 1066 ram only clocked @1000, a Radeon 5770 I play at 1920X1200 4AA everything maxed out and get an average of 38 fps and minimum 31.2. If you want proof I'll post a screen shot. This is just totally off though. you can see a huge difference in lighting shadows and even on the cars. If you need to upgrade an old card, then by all means get a dx 11 card. Dx 9 and 10 performance is much better and these cards use much less power and are very cool. On top of that what they don't talk about here is how much less cpu and memory load is with dx 11 over 9 far less.

 

[tburns1]

I think these new visual effects would probably be more appreciated in games that are not frantic in nature. You probably wont notice (or care) when you are blasting through an FPS. But for adventure type games, like maybe the upcoming Thief 4 or Bioshock 2, these effects would really add to the game play. I think the ambient occlusion would greatly add to these two titles in particular.

 

[hundredislandsboy]

This article is very misleading. First and foremost, the 5770 perform much higher than their showing in this article. Secondly, they used an outdated driver, 9.12 hotfix is the main one out and there's a big image difference between dx 11 and dx 9. Hardocp article is much more in depth, and is much better detailed than this shenanigan of an article. I have a core 2 quad Q9550 @ 4 ghz, 8Gb 1066 ram only clocked @1000, a Radeon 5770 I play at 1920X1200 4AA everything maxed out and get an average of 38 fps and minimum 31.2. If you want proof I'll post a screen shot. This is just totally off though. you can see a huge difference in lighting shadows and even on the cars. If you need to upgrade an old card, then by all means get a dx 11 card. Dx 9 and 10 performance is much better and these cards use much less power and are very cool. On top of that what they don't talk about here is how much less cpu and memory load is with dx 11 over 9 far less.

Dude I believe you, no need for screenshots. But. but, basically you're admitting to a serious bottleneck! A nice Q9550 OC'd to 4 Ghz and you're using a 4770?! What the mhz were you thinking?

 

[cleeve]

[citation][nom]t1daballer[/nom]This article is very misleading.[/citation]

It appears to me that your sensationalist comment is misleading, while the article is quite accurate.

1. First off, our testing was completed before 9.12 was released.

In any case, you suggest that the 9.12 hotfix driver makes a massive difference. According to AMD's release notes, the hotfix version of the driver "Hotfix that resolves the mouse being enlarged under Windows® 7, improves performance of OpenGL games, Provides CrossFireX™ support for Eyefinity & Provides support for OpenCL".

None of those things applies to DiRT 2, except crossfire support, and CrossFire was already supported in the 9.11 hotfix driver we used.


2. "you say "there's a big image difference between dx 11 and dx 9... you can see a huge difference in lighting shadows and even on the cars.".

Well, where is it? The HardOCP article you point out demonstrated the same image quality differences as this one from what I can see. Where is the "big difference in image quality"?

I heartily invite you to provide proof. Two screenshots - one of Dx9 and one of Dx11 showing the differences between the two would be nice. Good luck, because no review site has come up with that yet. "a huge difference in lighting shadows and even on the cars" doesn't exist. It's all in your head.

Are you basing this on side-by-side gameplay - like we demonstrate in our video - or wishful thinking and HardOCP's statement that "it is easy to tell a difference during normal gameplay".
While I have tremendous respect for the guys at HardOCP , I couldn't disagree with them more on this one. There is absolutely *NO WAY* to easily tell the difference between Dx9 and Dx11 during normal gameplay. All you have to do is click our video for the high-res version: there is no dramatic difference at all, very, very subtle differences maybe, but nothing you could possibly point out during an actual race when the scenery is flying by. And racing is "normal gameplay" in DiRT 2. It is not 'normal gameplay' to park your car on the side of the road and stare at rocks in order to see tiny and otherwise-imperceptible HDAO or shadow differences.


3. "I play at 1920X1200 4AA everything maxed out and get an average of 38 fps and minimum 31.2. "

And we got 25 min/30 avg. So what? Both results are very playable.

You pointed out yourself you're using a different driver, not to mention a million other variables we can't even begin to imagine. Did you even confirm that you're running the game in Dx 11 mode? It isn't reasonable to expect results to coincide perfectly when presenting results you got in isolation compared to a control group.





To summarize, you're complaining this article involves shenanigans when that's all you've offered. You demonstrate absolutely no proof of an appreciable increase in Dx 11 image quality whatsoever but you have no trouble calling the article down. You admit your test variables are different but then you complain that the article is misleading because you're getting different results?

Way to make a mountain out of a molehill. I'm not sure why anyone would attach their self esteem to the DiRT 2/Dx 11 cart, but it seems pretty clear you're not able to process this information objectively.

 

[hundredislandsboy]

It appears to me that your comment is misleading and the article is quite accurate.

1. First off, our testing was completed before 9.12 was released.

In any case, you suggest that the 9.12 hotfix driver makes a massive difference. According to AMD's release notes, the hotfix version of the driver "Hotfix that resolves the mouse being enlarged under Windows® 7, improves performance of OpenGL games, Provides CrossFireX™ support for Eyefinity & Provides support for OpenCL".

None of those things applies to DiRT 2, except crossfire support, and CrossFire was already supported in the 9.11 hotfix driver we used.

2. "you say "there's a big image difference between dx 11 and dx 9... you can see a huge difference in lighting shadows and even on the cars.".

Well, where is it? The HardOCP article you point out demonstrated the same image quality differences as this one from what I can see. Where is the "big difference in image quality"?

I heartily invite you to provide proof. Two screenshots - one of Dx9 and one of Dx11 showing the differences between the two would be nice. Good luck, because no review site has come up with that yet.

"a huge difference in lighting shadows and even on the cars" doesn't exist. It's all in your head.

3. "I play at 1920X1200 4AA everything maxed out and get an average of 38 fps and minimum 31.2. "

And we got 25 min/30 avg. So what? Both results are very playable.

You pointed out yourself you're using a different driver, not to mention whatever other variables we can't even begin to imagine. Did you even confirm that you're running the game in Dx 11 mode? not to mention a million other variables you have no idea of when presenting results you got in isolation compared to a control group.


To summarize, you're complaining this article involves shenanigans when that's all you've offered. You offer absolutely no proof of an appreciable increase in Dx 11 image quality whatsoever but you have no trouble calling the article down. You admit your test variables are different but then you complain that the article is misleading because you're getting different results?

Way to make a mountain out of a molehill. I'm not sure why anyone would attach their self esteem to the DiRT 2/Dx 11 cart, but it seems pretty clear you're not able to process this information objectively.

I think somebody just got owned! But don't hate! This is a good thing. You can't come on to TH representing personal opinions as facts. Empirical evidence is the name of the game here. lol.

 

[cleeve]

[EDIT] Ooops! I responded to the wrong person, my bad.

 

[SchizoFrog]

There are going to be a lot of 'heters' out there with regards to articles like this as it proves that all those eager people pretty much wasted their money (at least for the time being) by rushing out and buying a 5xxx series ATi GPU because of its supposed DX11 support. Yes, they are good cards but are they worth the £'s??? Once again I think not and I am happy to wait for my next upgrade.

As for DiRT 2... While it has killed off the serious rally element of the game, I have to admit that I have so far found the game rather enjoyable.
There seems to be no point in buying other cars unless you need one to enter a race as your first cars seem to be better than the others on offer, so whats the point in buying them? The focus on dashboard decorations, liveries and other such rubbish is a little annoying. But as a pic up and play game, it succeeds... could be better, but definately worth picking up.

 

[envolva]

[citation][nom]schizofrog[/nom]There are going to be a lot of 'heters' out there with regards to articles like this as it proves that all those eager people pretty much wasted their money (at least for the time being) by rushing out and buying a 5xxx series ATi GPU because of its supposed DX11 support. Yes, they are good cards but are they worth the £'s??? Once again I think not and I am happy to wait for my next upgrade.[/citation]
If you don't need a new graphic card why would you rush to buy a new graphic card? You know they will be cheaper in a while, or you can wait for future releases. So your initial comment doesn't make much sense to me.

If you do need a new graphic cards, are you wasting money buying an ATI 58xx? I don't think so. Great performance with old games, low power, and as a bonus you are ready to DX11.

Would you buy a NVIDIA card with about the same performance, same price power hungry, and without DX11 support?

 

[Guest]

ATI just finished their newest super gaming model,and already you are limited to enjoy the fullness of graphic detail!

I wished something could be done on the power requirements of those higher end cards.
Like have a 60W in power graphics card, and be able to play almost any game at almost any resolution with almost any detail to the max on 60fps.

 

[fulle]

Dirt 2 came with my Asus branded 5850 (its just a reference card) via a steam activation code.

I'm not a console gamer, so I was pretty much instantly annoyed. After you get signed up for windows live, and all that, the first thing you'll notice is that you can't change your graphics settings at the main menu screen! You have to create a whole profile, and name your player, and so on before it'll let you check to make sure everything's setup properly on the graphics side of things. At that point you'll notice the menus are still console style, making you pan wide left and right, and only showing one option at a time. So, I check it out and it defaulted to max settings at the correct res... so I start a race. TOTAL MISTAKE trying to do this with a keyboard and mouse. The controls are atrocious! So, I plug in a 360 controller... and it hits me... they're playing freaking EMO MUSIC. KILL ME NOW. Deleted.

Still waiting on a DX11 title that's not a bad console port.

 

[SchizoFrog]

[citation][nom]fulle[/nom]Dirt 2 came with my Asus branded 5850 (its just a reference card) via a steam activation code.I'm not a console gamer, so I was pretty much instantly annoyed. After you get signed up for windows live, and all that, the first thing you'll notice is that you can't change your graphics settings at the main menu screen! You have to create a whole profile, and name your player, and so on before it'll let you check to make sure everything's setup properly on the graphics side of things. At that point you'll notice the menus are still console style, making you pan wide left and right, and only showing one option at a time. So, I check it out and it defaulted to max settings at the correct res... so I start a race. TOTAL MISTAKE trying to do this with a keyboard and mouse. The controls are atrocious! So, I plug in a 360 controller... and it hits me... they're playing freaking EMO MUSIC. KILL ME NOW. Deleted.Still waiting on a DX11 title that's not a bad console port.[/citation]

What the hell are you on about? If you had a 360 controller why would you even try to play a racing game with keyboard and mouse? As for the music games like this always feature a selection of popular chart music and the current charts are riddles with this music and most of it is decidedly NOT EMO, so please don't make wild generalisations. As for the graphics settings, so what if they are tucked away a little, every setting was as you said set to max, as they were for my 9600GT and as I only mess with settings once, if that, again I'll say, so what if they are tucked away?
This is not a BAD console port, its actually a very good one. If you are waiting for a true DX11 game then happy waiting but it will be years, when and if the console market moves to DX11 with their next gen consoles, apart from an exclusive like Crysis was for DX10 but Crytek won't make THAT mistake again, they have already stated that their next engine will run on PS3.

 

[SchizoFrog]

[citation][nom]envolva[/nom]If you don't need a new graphic card why would you rush to buy a new graphic card? You know they will be cheaper in a while, or you can wait for future releases. So your initial comment doesn't make much sense to me.If you do need a new graphic cards, are you wasting money buying an ATI 58xx? I don't think so. Great performance with old games, low power, and as a bonus you are ready to DX11.Would you buy a NVIDIA card with about the same performance, same price power hungry, and without DX11 support?[/citation]

Maybe you misunderstood me, not sure.
The point is DO YOU NEED TO UPGRADE YOUR GPU? Or are you just buying the latest GPU because of the hype? If I had to buy a card right now, say due to a failure then yes, I would seriously look at the ATi range, but for now I will wait until I believe that there is an general benefit to having DX11 hardware (which isn't right now) and I can't play the games I want to on the settings I want to with my current hardware. I would also wait until nVidia comes to the fight as at least that should bring prices down further.

So, my comments were about those who would and HAVE upgraded their GPUs purely because these are the latest cards on the market.

If you can play the same games, at the same settings on a £100-£150 card as £200-£300 cards then YES, you are wasting your money, now matter if it is nVidia OR ATi.

 

[notty22]

I've played driving games forever, outside going to the arcade , this is a very fun game.
Least favorite part is the monster trucks, favorite is the Rally cars.
DX11 is a non factor. Don't hate the game because dx11 implementation slows fps to a crawl. lol-----That's ATI's fault.

 

[redgarl]

LOL... do anyone really see a difference?

What is sad is that we are going to see these kind of result until a game is fully developed in DX11... not just a stupid DX9.0 port made for an outdated technology called the xbox 360 which is 3 generation of DX late.

 

[redgarl]

[citation][nom]notty22[/nom]I've played driving games forever, outside going to the arcade , this is a very fun game.Least favorite part is the monster trucks, favorite is the Rally cars.DX11 is a non factor. Don't hate the game because dx11 implementation slows fps to a crawl. lol-----That's ATI's fault.[/citation]


LOL... of course.... ATI fault... let be serious for a minute... you are actually trying to tell me that a DX9.0 title made for the xbox 360 run bad in DX11 because of ATI?

Really good logic, Ati is not the one who developed this game. Also this wasn't meant to be a DX11 showcase in anyways.

 

[agawtrip]

2004 - half-life 2 --> "infant" dx9
2007 - crysis --> full dx9

2007 - http://www.gamespot.com/features/6174799/index.html
- mostly dx9 ported to dx10
2009 - stormrise / shattered horizon --> full dx10 / 10.1

2009 - dirt 2 --> "infant" dx11

if the trends continues, it will be 2 years before we can see games using the full capabilitites of dx11.

 

[SchizoFrog]

[citation][nom]agawtrip[/nom]2004 - half-life 2 --> "infant" dx92007 - crysis --> full dx92007 - http://www.gamespot.com/features/6174799/index.html - mostly dx9 ported to dx102009 - stormrise / shattered horizon --> full dx10 / 10.12009 - dirt 2 --> "infant" dx11if the trends continues, it will be 2 years before we can see games using the full capabilitites of dx11.[/citation]

Not sure I fully agree with you for a couple reasons.
Stormrise was a complete mess and non event as far as releases go and I think most people have either never played or don't want to think about it anymore. As for Shattered Horizon, It has things going for it but I can't help but feel that it was more of an application experiment for their benchmark engine. So we are really still waiting for DX10 games. Another factor would cbe that todays consoles couldn't get close to running DX10 let alone DX11 so we will have to wait until the next gen consoles come out before developers even think about moving things on. We may get the odd game with a few tweaks and extras (that no one will notice most of the time) and maybe an exclusive or two but that is it.
With Microsoft's announcement that Windows 8 will hit in 3 years time, and that is around the same time that the next gen console will hit (if at all) then I can easily see that DX11 will NEVER be fully supported by developers in terms of games and eventually will just be skipped for DX12.

 

[megamanx00]

Kind of wanted to see the difference between DX10 and DX11. After all, most gamers have DX10 cards by now.

 

[liquidsnake718]

Well for those who think getting the 58xx series is a waste, its not, it debuted with a reasonable price point and those cards are a rare specimen that has proven its worth. At a great price point it has made me change from my old Nvidia cards(9800gtx+)to a 5850. It shows as my last 5 cards were all Nvidia cards. Starting from the voodoo fx, to the 9800gtx+.

I know that DX11 is still new and next year we will see more games coming out for it. I remember when DX10 and 10.1came out, only a few games like Assassins Creed and Crysis had DX10 support. This time Microsoft was busy with Windows 7 launch, so I guess DX11 will be the focus on the graphic front this coming year.

 

[Guest]

I must say this is probably one of the best looking racing games I've seen by far. Add onto that its very easy on the CPU while utilizing multiple cores very well. I wonder if any later patches will improve DX11 performance?

 

[Guest]

All that you put on the table from nvidia is a GTX260? WTH happened with GTX285 or GTX295? You test 9 products from ATI and 4 from Nvidia? With what purpose?

 

[zool_50]

Anyway to add 4850 in the charts ?
The new 5700 series are quite disappointment. The price/performance is just tragic. How can they expect for me as a 4850 owner to downgrade to 5700 that costs more, it has almost same performance and with dx 11 enabled the fps sucks hard. And to the price, it has 40nm gpu and 128bit pcb.
The only real new cards are 5800 series which are priced much higher than the old 4800 series at lunch. Its quite sad that with a new series the performance downgrades for the same costs.
Lets hope fermi will beat 5800 to dust and the prices will be much more realistic.