Friday update and discussions 2011-03-18

[Blackdog_kt]

Originally Posted by recoilfx

While I agree that Maddox may have made some wrong choices regarding graphics development (IE, cloud & water should have been concentrated more vs the diminishing returns on certain nitpicking details), but given that they are such a small team with finite resources and given niche market share of DX11(http://store.steampowered.com/hwsurvey), not shooting for DX11 is understandable and wise. Even Crysis 2 is DX10 upon release!

It's wrong to assume that the engine will be antiquated upon release because it doesn't support DX11 - it's only a rendering path. You don't need to replace the whole engine (FM, DM, atmosphere modeling & resources management, blah blah blah) just to get DX11.

I'm going to go ahead and agree with this. The percentage of people with DX11 capable cards among the total gamer demographic is still somewhat small to warrant delaying the game a further 3 months or so just to implement DX11 support.

I mean, if Steam's surveys say so (ie, people who usually play more GPU-intensive games than the rest of us) it certainly is not something that's so widespread yet that would be seen as a terrible oversight. It's almost 95% of the people in that survey that don't have a DX11 capable GPU, most of them won't even bat an eyelid if a title doesn't support DX11 because they can't use it anyway, unless they are planning to upgrade very soon that is.

In any case, let's assume an optimistic 20% will have DX11 capable GPUs during the next 3 months and showing signs of further growth...that would be a good point to really start working on it. Until then, i'm happy they stayed focused on how the aircraft are flown and operated. Sure, this is subjective and it's only my opinion, it's just brought about by the simple fact that CoD still is a simulator game about WWII aircraft and not a tech-demo for using the biggest amount of current day technology and APIs.

Unless DX11 support can somehow take a substantial part of the processing load off of the CPU, i don't really see a major advantage. I think it can (at least that's what google translator seemed to mean when i read that German interview), but this opens the can of worms that stands at the center of the issue: we can't have people flying with different FM/DM and different fidelity in physics, depending on their GPUs.

This means that even if DX11 can transfer some of the CPU's load to the GPU, we'll still have to wait until almost everyone has a DX11 GPU before off-loading critical modules of the sim there. Until then, even when it's supported by the sim it will most probably be used for stuff that's optional and as such, non critical: visual detail settings, maybe a revamped sound engine and so on, but not for crunching the math that drives the core of the sim because that would mean forcing people with older equipment into a lower difficulty bracket by default. I'm not saying it won't happen at some point (having the GPU calculating FM/DM parameters), if the new series has as long a life as the previous one it will certainly be done...probably by the time that DX11 cards are as common as DX9/10 cards are today.