Finally decided to see if there was an end to this story and most of it is covered already. :-) I was suspicious of any performance enhancing bug that did not affect image quality and is good to know is actually legitimate because you can just reuse the depth buffer in DX10.1 instead of second pass.

I found the Rage3D link was very enlightening as well:
Did we find the glitches that everyone has been talking about when making reference to the 10.1 path? In a way - we do have that missing dust that qualifies for that category, albeit we don't know yet if it's simply a driver bug or something wrong with the pathway itself. Other than that, there's the more intense lighting, but that actually seems to show that there's a bug with the DX10 pathway, given the fact that DX9 has the same, more intense, lighting as 10.1, and UBi said that 9 and 10 (and by extension 10.1) should be nearly identical visually (shadowing is different between versions).

Lets remove the DX10 path too? :-)

I was never a strong follower of either camp (just choosing the better GPUs that are released), but now I'm thinking of shoving nVidia the finger and going to get myself some of that red stuff.

No money changed hands? Only the fools will be convinced. Ever heard of value-in-kind sponsorships? No cash may be transferred, but those kind of sponsors reap benefits in the 6 figure ballpark at least!

As for Ubisoft:

Gamer: DX10.1 gives better gaming experience
Ubi: let's remove DX10.1 and lower gaming quality while claiming we're delivering higher quality.
Gamer: Why remove?
Ubi: Cos it's buggy.
Gamer: How is it buggy?
Ubi: It just is cos nVidia said so.
Gamer: Does this affect gaming performance?
Ubi: We are committed to ensure gamers are given high quality performance in games (but will remove DX10.1 anyway so stfu).

Shame!

Still up in the air is the question of whether Ubisoft will restore DX10.1 support to Assassin's Creed in a future patch.

http://www.pcgameshardware.de/aid,645430/News/Ubisoft_No_Direct_3D_...

. . . pushing common standards like DX10.1 . . .

Thanks. Thanks for crapping on the Linux, Mac, and BSD crowds. If you want a "common" standard, stick to OpenGL, yeah?

Actually, speaking as a member of the Linux crowd, I found this article rather humorous -- all this uproar over one patch for a single game -- is this actually important to people? It's rather otherworldly -- though I suppose if I was addicted to buying $600 graphics cards, I would want to be getting my money's worth. And for $600, I should hope that means nVidia, not ATI!

Wow, just wow. Evil. TR hinted at removing TWIMTBP games from reviews and I think that is genius.

So, they had time and money to work on this DX10.1 implementation and get it working well, and then suddenly can't find the time and money to do a little polishing of it.

Granted, I don't know if they hit a dead end or some such, but... Sorry, I don't believe one friggin' word of it..

Just because no money have been exchanged, doesn't mean Ubisoft don't find the cooperation itself valuable enough.

until Nvidia also has a DX10.1-capable GPU on the market.

If UBI does reintroduce DX10.1 into this game, then it'll be just around the time Nvidia introduces their DX10.1 cards. Don't believe me? Give it time.

This is why my PC's have AMD CPU'S and graphics cards.

This is a great story however, I believe that an interview with ATI's developer relations team would cast some light on the situation. We have Ubi and Nvidia's take on the situation AC's DX10.1. Lets hear what ATI developer relations team has to say about this. For example:
-Are they having difficulty with Ubi to improve on DX10.1 code in AC?
-Are the problems found a driver issue which can be fixed in a future Cat release?
Or
-Are they in fact currently working with Ubi to improve on DX10.1 code for a future release
-etc.

Questions like this would shed light on the situation IMO.

what's so hard to undestend? Xbox has exclusive games like Halo, and PS 3 metal gear etc..
Nvidia is triying to make more out of the title and just like they did with Crysis they paid Ubisoft to optimaze the code for there cards, but is more simple to uses DX10.1 then 10 so they did so 10.1 (or maybe where paid by ATI :))
Will implement DX10.1 in a future patch?
this question shoud have be put to Nvidia , yes is 9900 series will have 10.1 else no. clear & simple.
[There where money involved?]
YES there are , why ? evry thing is about money,adivertisement is money, etc ,
PS TR shoud put AC in Damage Lab (with DX10.1) .for testing all the new and old cards .

Why not just add the option to switch from a DX10 to DX10.1 renderer? It would work kind of like how you can select a DX9 or DX10 mode in some games (CoH). Is the too difficult to implement that technically?

Game developers should never cater to one manufacturer or another like this. They should have given the end user a choice.

It's disappointing that there is never a peep from any sort of ATI dev rel.

Thanks for the update, Scott.

Even if they're all telling the truth, it still doesn't explain why Ubisoft decided to completely remove the 10.1 path, rather than just putting in the option to disable it. I suppose that just comes down to time and money. Whatever Nvidia's role in all of this, Ubisoft doesn't come out of it looking particularly good, especially given that the high system requirements point to a really lazy port. I'll leave the rant about cross-platform games for another day.

Who cares if "no money has changed hands". Money isn't the only form of payola. Nvidia says "we'll promote the game if you drop DX10.1" Ubi says, "sure, no problem." There you go, no money changed hands, but still pure evil. That's my flamboyant speculation at least.

This is why TR kicks everyone else's ass.

Moreover, has anyone actually played this game?

Silus, I am fixed on the fact that its "plausible" to fix the problems they believe exist not remove DX10.1 altogether. The AF issue was fixed in a driver hotfix. The white dots that you referred to could be a driver issue or game bug. What is "not clear" is how those white dots are produced. However, what is believable is that AMD/ATI's developer relations team are ready to work with Ubi with fixing Assassin's Creed's DX10.1, not remove DX10.1.

Update 5/9/08:

/proceeds to praise TR in a bubbly manner

Silus, per the article it's not hard to believe that
AMD's developer relations team are ready to work with Ubit with fixing Assassin's Creed's DX10.1. This alone makes it simple and clear. There is no reason to remove DX10.1. BTW, of course DX10.1 removes the render pass! That's what 10.1 does. The same effects in 10.1 take 1 pass, in DX10 it takes more then 1 pass. The additional rendering path in DX10 is redundancy not accuracy. I hope you are not implying it takes excessive passes (overhead) to properly render post process effect in AC using DX10.1.

Silus #62
Bugs no less. It could be a ATI driver issue or a bug in the game itself.

Kraft #75
Not only do they provide tools, but also:
-"Support Team"
-Game Test Lab
and other advice on how to code the game. The article clearly states:
-they help developers code games
-has over 100 people working on programme
-assigns engineers their support team to work "in-house" with development teams
-etc
It might be useful to read the article again. TWIMTBP program offers more then just "a slew of useful tools" in exchange for advertisement.
Once you re-read the whole article again and understand it I won't hold it against you :)

TR: What other image quality and/or performance enchancements does the DX10.1 code path in the game offer?

Beauchemin: There is no visual difference for the gamer. Only the performance is affected.

TR: What specific factors led to DX10.1 support's removal in patch 1?

Beauchemin: Our DX10.1 implementation was not properly done and we didn't want the users with Vista SP1 and DX10.1-enabled cards to have a bad gaming experience.

Improve performance = Bad gaming experience

more like

Improve performance on AMD = Bad experience for Nvidia.

Kraft75 #33
TWIMTBP article I linked you to clearly states that they help remove bugs, optimize and test games. This is far more then slap-on advertising. I suggest you read the entire article again as you are not comprehending it correctly as the portion I linked you to clearly states "More than just marketing"

Silus #39
That review regarding lowered AF was a result of a bug in the Cat driver that was fixed in the same update used for improving Vantage performance. [H] has not revisited the AF problem using the updated drivers which corrects the AF problem. Also, it's obvious the game has bugs. However, it's not a result of post processing.

Interesting article:
http://www.pcgameshardware.de/aid,641035/Test/Assassins_Creed_D3D_1...

While I appreciate the effort you guys put into this I am not exactly as bubbly about the conversation with Mr. Beauchemin as others.

The questions and answers summed up what I have already read. Which in of itself is nice to get a final word on, still I would have liked some harder hitting questions like:

"Wouldn't a option to enable/disable this be prudent instead of pulling it out completely?"

"Did Nvidia's TWIMTBP have any impact on the decision to pull the feature?"

We are talking about a rather substantial difference in speed over some visual differences. *adjusts tinfoil hat* His response of: "We are currently investigating this situation." is pretty weak.

Quack.

You'd think nvidia would have learnt that the community backlash from being a$$holes will hurt its bottom line.

I went with ati on the Radeon 9700 Pro instead of nvidia because of their cheating shenanigans with drivers.

Now that they're doing the same with developers, I'll probably go R700 instead of G200 for my next upgrade.

EDIT: Oh, and Ubisoft bowing to this bull$hit? Don't expect me to buy any games from you for a while, either.

This confirms my suspicions when I first heard of the DX10.1 controversy with Assassin's Creed.

Ubisoft does not have the balls to make Nvidia look bad.

Nvidia, here is a hint. Stop releasing the same darn architect. It is almost two years old! Die-shrinks are fine and some tweaking. There comes a point where you have to move on though. Just learn from ATI's own mistakes of trying to hold out with R3xx architect.

This is coming straight out of a Intel/Microsoft play book.

Putting pressure on companies / cheating to get ahead.

Is it any wonder my current quad core is AMD and my next video card purchase will be ATI.

And yes, I give a crap about getting the most for me money, but I also prefer to support companies that do things right.

I can live with the few FPS I lose better then I can live with this kinda crap.

Great work, Scott!!!

What a pity I've just bought a 8800GTS...

nvidia is satan. nuf said. :)

So when its played on NVidia cards there's some bugs apparently if the DX10.1 code is running but on the ATI cards it runs well and faster?
Given the nonchalant way that this feature was dismissed by the "graphics gurus" here...where does that leave them now?
Congrats to TR for investing time and effort on this initial investigation.

Not convinced with the way, this guy has answered the questions. Also, some more questions were need to be asked.. Oh btw the last answer was full of bull...

But still good work sir..

Has any game shown anything remarkable with DX10 verus 9 alone?