Sign in with
Sign up | Sign in
Your question

DX 10 has perverted the trend for R&D financing

Last response: in Graphics & Displays
Share
June 5, 2007 12:25:11 PM

Having now sampled some DX 10 hardware and software (patched CoH and DX10 codesamples) I am beginning to come to the conclusion that the current strategy for MS and GFX makers has done rather well in generating signficant revenue for a technology that is basically a paper launch.

To elobrate it would appear none of the current generation of so call DX 10 hardware can run DX10 with anywhere near required performance (8800 and 2900 excl'd, and even their performnance is far from stellar), this could change with driver/software optimization but to get a 30% aggregated performance improvement across the generation seems unlikely. Rough educated guess on performance improvement requred. And also the next generation will probably have to fit in with the corporations speed bump strategy ie. geforce 9 can not cull to much off the remaining 8 series sales

And with this state of affairs developers are unlikely to invest much resources in to developing DX10 code paths when in reality most people will have to resort to the DX9 for acceptable eye candy/performance ratio. There will probably be noticeably exceptions with games that have cutting edge tech as one of their unique selling points, but thats like 1 or 2 games a year.

People will cite immature code and the likes but even improvements here would seem unlikely to produce signifcant enough results. But just to help shoot down my own argument I will be trying the nvidia beta drivers when i get home from work to see if there is any change, nvidia did recommend these for CoH which i only just found out. Also coders can do some wonderfull things with optimisation, a'la late PS2 games, but its the poor performance DX10 samples thats is worrying me because these are simple enough techniques being demo'd
June 5, 2007 12:59:50 PM

awhh ur still whining about ur 8500gt and dx10 :roll:
June 5, 2007 1:05:08 PM

Nope just DX 10 and its associated hardware in general. Don't think i mentioned the 8500 and wouldn't expect anyone to be interested in my previous posts :) 

I like the card as it ends up for what its good at, dx 9, but it has got me thinking about the state of affairs with DX10 and i can't help but think 1 year to soon.
Related resources
June 5, 2007 1:16:22 PM

well what did u expect? getting higher fps in dx10 than dx9 when u patched it? ..
i dont really care for dx10 yet and specially not vista with its DRM crap but we will see in a year or so :) 
June 5, 2007 1:33:17 PM

I just wasn't expecting the large GPU overhead for DX 10 as is currently. Its not to say that it is completely broke or anything. But it feels like it was released a hardware generation to soon and the corprats are just milking the consumer for a phantom tech thats is at least a year away from realisation.

Oh the DRM debate, off topic i suppose, but i have yet to notice vista's DRM mechanisms, and have my fair share of *cough* ripped/backed up content. I have no HD content which is probably where u'd notice the DRM though.
June 5, 2007 4:01:13 PM

It will be a while before programmers and game designers get associated well enough with dx10 to produce eye candy that really shows DX10 ability anyways, so you probably won't see a huge difference between dx9 and dx10.
June 5, 2007 5:55:45 PM

Definitly true gold, i have a few myself for geometry shaders and access to the vertex buffer, creating poly's on the fly great stuff.

Anyway its not even so much extra eye candy as a level equivalent to admittedly higly optimised dx 9 code, which seems unlikely though i'm about to test the nvidia beta driver and if that gives even a modest boost to dx10 performance i may start eating my words. About dinner time here anyway :) 

I just hope they can find enough shader coders to make loads of dx10 games, I really should get my finger out :p  Though i think i'd need dx10 delayed by about 3 years to get anywhere near good enough :( 
June 5, 2007 6:17:58 PM

Oh FFS. You haven't even seen a native DX10 game yet... you're basing your conclusions on some half baked DX10 demos and patches. Also, I have to wonder if you've seen the latest DX10 nvidia demo in action: Human Head

I think you'll find that when Crysis and Alan Wake is released you'll be regreting this post. DX10 is biggest change in API's ever for developers.

I won't argue that M$ attempting to use DX10 to sell Vista, but lumping nVidia and AMD into the "conspiracy" is absurd. Read all the comments every game programmer has to say about DX10 and then come back in here and comment. They all love it.
June 5, 2007 6:51:18 PM

Jesse to clarify about the demo's i'm talking they are the ones that come with the SDK to show off new features one program at a time not something trying to shove every bell and whistle in, and they most certainly don't run that well at all.

I am not saying dx10 is shit or that its some conspiracy other than the fact corps will do everything in they're power to squeeze out a dollar. Nor am i getting in to a debate about the vista tie in. Indeed i welcome the changes in dx10 it does give a sense of more togetherness in the API

All I really wanted to highlight is that the current generation (as a whole) IMO will not cut signifcant mustard in DX10, the 8800's are fine but in a short while the 320 gts could be consider entry lvl dx10. And since 8800 owners are most certainly in the minority than means no dx10 market for a while i believe.

As for the games Alan Wake is way sometime in 2008 and Cyrsis can afford to ignore the diminished returns of lack of dx10 support, only 8800 and 2900 owners are going to be playing dx10 Cyrsis, because they emphasis tech in they're games and plus i wouldn't be surprised if they were getting kick backs from MS.

It could be said MS pushed DX10 to soon or that video card makers push a tech that on the whole they knew wouldn't cut it but figured no software so what the hell.

I'm just a bit miffed i suppose, i want my new tech and i want it now :) 
June 5, 2007 7:23:50 PM

I speak for myself here when I say, i have an 8800GTX and Windows Vista.

I do have COH and have played the DX10 version needless to say its insignificant to warrant a change in Graphic cards and coding.

What did Relic do? that makes DX10 COH so much better then DX9, ummm short grass? i mean wtf you barely see grass from such a high view anyhow, "litter objects", besides seeing no trash on streets(pun intended) all i see is "rocks" and "Flowers" amazing :roll: .

Now i will stop bashing relic because they just patched to DX10 and didn't build the game from the ground up on it. Right now DX10 is still young yet many continue to bash it. its been what 5 months? DX9.0 has a whole 3-4 YEARS to mature and we have seen many revisions, 9.0B/C. I personally like Vista and choose it over XP because of the features and what i say., ease of access. Though I'm a big gamer and many games are still unstable (BF 2, MTW2, RTW) it is because they were not designed on Vista but on XP and on DX9.0.

We also have to remember these Graphics cards are the first generation of DX10 with 8800GTX/GTS out even before DX10 was and ATI coming very late to DX10 party and with only one version HD 2900XT 512MB
The architecture is still young and with the G92 supposedly coming out in X-mas and maybe :wink: HD 2900XTX we'll see if it improves or not.

DX10 has many great possibilities of course at such an early stage everyone is still learning to use it and later on we shall see much more efficient games and here is hoping SLI drivers that work. DX10 is going to become mainstream for gamers whether you like it or not all future games are still going to support DX9.0(much like how HL-2 supports DX8.0) but are going to be DX10 supported and touted. MS understands that gamers make up about roughly 60% of the market and selling Vista with High-end "Vista ready/certified parts" mean $_$. Already all pre-built computers come with Vista good luck finding a new pre-built computers with XP installed.

People are either resistant to change or fiercely resistant to it. new technology is not always embraced with love, one such example is Physx processor, which is you haven't noticed is specifically touted towards gamers.. not toward the average person using MS word....though Advanced Physics in MS word would be cool.... :lol:  , anyway Love it, hate it, Vista is here to stay and conquer the computer world, one computer at a time. Each of you nay sayers will slowly move over to Vista, i mean is there one of us who still use Windows ME, or 2000? but Vista is not like those which were basically redesigned Windows 98's that came with nothing.....

This post was brought to you by Microsoft(TM) and Pepsi :twisted:

Xazax
June 5, 2007 7:33:08 PM

Quote:
Also, I have to wonder if you've seen the latest DX10 nvidia demo in action: Human Head


Read the requirements for Human Head, they list Windows XP as a target OS. In other words, Human Head is a DirectX 9.0c demo. Sorry to burst your bubble on that one. It's still impressive, though.

In regards to DX10 performance, keep in mind that the only things we've seen so far have been Company of Heroes (a game that can make systems cry even in DX9), Lost Planet (a console port with DX10 instructions thrown in for good measure), and Call of Juarez (a game originally designed for DX9 using a little-known engine). They really don't give us a lot to talk about yet... although we all know Crysis and UT3 are just around the corner.

Also, I wouldn't exactly call an 8800GTS "entry-level" DX10... especially when overclocked (the 320MB version I have I've OC'd stable to a relatively stratospheric 650/1520/1860)... I think both ATI and nVidia have at least some idea of what they're doing with their GPU design, and that their cards with the gaudy numbers in the hundreds place (8s and 9s, for those of you who went to public schools... like me) should handle at least the first batch or two of DX10 games without much of a hassle.
June 5, 2007 7:37:25 PM

Hehe nicely put and true of course, but thats where the title comes in in my opinion the process should go

investment > R&D > Product > Profits > payback and reinvestment > R&D....

but its currently

R&D > Promise of product > Profits > more R&D and fat chairman cheque :)  > Product > Profit

to many profits there for my liking.

Anyway way the crux i think is that DX10 is a migration thats is argueably worse than any that came before in terms off its complexity ie u need two completely different code paths where as before u could just opt to not run that new fangled sm3.0 program. And because of this i think the transitional period needed some strong hardware not a generation that hints at the new tech but can never truely deliver.
June 5, 2007 9:50:03 PM

Quote:
Also, I have to wonder if you've seen the latest DX10 nvidia demo in action: Human Head


Read the requirements for Human Head, they list Windows XP as a target OS. In other words, Human Head is a DirectX 9.0c demo. Sorry to burst your bubble on that one. It's still impressive, though.

In regards to DX10 performance, keep in mind that the only things we've seen so far have been Company of Heroes (a game that can make systems cry even in DX9), Lost Planet (a console port with DX10 instructions thrown in for good measure), and Call of Juarez (a game originally designed for DX9 using a little-known engine). They really don't give us a lot to talk about yet... although we all know Crysis and UT3 are just around the corner.

Also, I wouldn't exactly call an 8800GTS "entry-level" DX10... especially when overclocked (the 320MB version I have I've OC'd stable to a relatively stratospheric 650/1520/1860)... I think both ATI and nVidia have at least some idea of what they're doing with their GPU design, and that their cards with the gaudy numbers in the hundreds place (8s and 9s, for those of you who went to public schools... like me) should handle at least the first batch or two of DX10 games without much of a hassle.

Indeed.
June 6, 2007 12:51:12 AM

Quote:
I think you'll find that when Crysis and Alan Wake is released you'll be regreting this post. DX10 is biggest change in API's ever for developers.
....

Read all the comments every game programmer has to say about DX10 and then come back in here and comment. They all love it.


Of course, developers and programmers love DX10 because they virtually wrote it along with MS.

The OP appears to demand that as soon as a new API is written, the ultimate hardware and drivers must appear immediately, and all games in development now, those about to be released and those already released, must be DX10 capable, fully implemented and deliver 4 x performance and quality all at once.

DX10 is a massive step forward as you've said, and it will take time to fully mature. The fact that it is tied in with Vista has made it an easier target for the conspiracy theorists and doomsayers out there.
June 6, 2007 1:08:33 AM

Yeah Abyss I would love all those things :)  But i can forgive bad early drivers and don't expect the software really until the hardware is there and thats the issue i don't think its there. It having no chance of getting anywhere near critical mass for a while without either a intermediate refresh of geforce 8 or the release of 9 mid-range.

Maybr its just been overhyped to much to soon and thats why this generation of hardware leaves a bad taste.
June 6, 2007 1:32:22 AM

Quote:
Yeah Abyss I would love all those things :)  But i can forgive bad early drivers and don't expect the software really until the hardware is there and thats the issue i don't think its there. It having no chance of getting anywhere near critical mass for a while without either a intermediate refresh of geforce 8 or the release of 9 mid-range.

Maybr its just been overhyped to much to soon and thats why this generation of hardware leaves a bad taste.


I'm sorry, but your expectations are just not realistic.

The hardware cannot lead the software in this case - the API is introduced in conjunction with developers, and it is finalised (and will be update dof course over time). Hardware takes longer to develop and manufacture, and the drivers for it take longer still.

Do not forget that the changes in DX10 over DX9 are far bigger than any previous changes made.

The point that there is no mid-market is also not really relevant without understanding how the PC industry works, and I do not intend to go into a detailed discussion on that.

Putting it simply, I'll use Nvidia as an example purely because their cards are out and I can't be bothered to typed 8xxx / HD2xxx every time.

Nvidia have spent a lot of money developing their 8xxx cards. They want to sell them to maximise their ROI. The marketing and support of their 7xxx cards is now dropped down. All 3rd party PC builds are now 8xxx cards because the manufacturer supply of 7xxx cards, whilst not ceased, is dwindling.

Thus the market is being driven. Whether or not you as an upgrader feel the 8xxx cards are good value for money is pretty much irrelevant. You can moan about it here until the cows come home, but as you like me represent a small part of the market, we're not driving it but at least we are on the train. This train is the critical mass, and if you're an upgrader, don't be fooled and think you can change it's direction.

The change will come about in time, with better drivers, better revisions, better development.

If you're a system builder, given the choice you'll put a 8600GT in your system instead of a 7600GT. You can put the nice DX10 sticker on the case, and the people at EA will be happy because you'll take notice of the lovely DX10 screenshots on the box cover of Crysis. Fortunately, for the OEM market, the cost isn't much different, and if it is more expensive, it's offset because more people are buying because they are 'upgrading'.

Sure, if you want to moan about there being no performance difference between your 7600 and a new 8600, fine. But what are you going to use it on? Hardware development doesn't coincide with software development which is out of sync with programming development which doesn't quite fit in with our buying habits etc etc. If you buy a 8600 today, you won't be disaapointed with DX10 performance because you can barely get it anyway.

I'm starting to waffle, doubtless induced by the large amount of Magners I have enjoyed tonight, but please bear in mind that there are always steps on the continued way to progression and improvement, and some of them will be different sizes at different times. Development is so fast today it will change in a few months,so if you think a small step is worth moaning about today, meh.
a b U Graphics card
June 6, 2007 3:12:21 AM

It's hard to get excited one way or the other over something that just isn't there. It's a catch 22.

Quote:
The OP appears to demand that as soon as a new API is written, the ultimate hardware and drivers must appear immediately, and all games in development now, those about to be released and those already released, must be DX10 capable, fully implemented and deliver 4 x performance and quality all at once.


I think you over exagerate the OP's expectations. Spoken like a hardened techie, I think your geek is showing.

I have to go with the OP on this one and play the role of Joe Average consumer, to "demand" that new hardware be released with at least one solid driver set and a few titles to showcase the new eye candy along with a noticeable performance increase is what sells product and makes the cash register ring. The consumer drives the market, not the other way around.

I'm making a sweeping generalization and referring to the tech industry as a whole; the big players in the tech industry (Intel, AMD, M$, nVidia, etc) seemingly push new tech and "standards" on an unwilling consumer based on a "demand" for the next big thing. Anyone who has ever lived knows that they (hardware and software developers and companies) are just peices to the puzzles and it takes time for them all to line up and fit. But as of now, today, DX10 is the latest in a long list of "technology improvement" debacles that has left Joe Average consumer scratching his head thinking, "WTF?!"
June 6, 2007 3:38:23 AM

Quote:
Do not forget that the changes in DX10 over DX9 are far bigger than any previous changes made.


This is true. You really can't compare the DX9 launch with the DX10 launch since there are so many differences. For one DX9 was developed for 98 and XP. DX10 was developed for Vista only to promote the sales of it. Not only that DX10 is supposed to be a big jump from DX9. Whereas DX9 was very similar to DX8 and DX7 for that matter.

Add to the list that MS really pushed this thing and made it sound like it was the best thing since sliced bread. It may be a great improvement but it will take a while. Because of DX10's immaturity, maybe the Nvidia 8800 and ATI 2900 may not be so future proof after all.
June 6, 2007 3:54:20 AM

Current DX10 hardware probably wont run DX10 games on maximum, but I'm sure they'll do a great job at it. Crysis has shown an official DX10 demo running on "the G80 GPU technology" at max graphics. It wasn't smooth(seemed to be 20-30fps) and probably SLi, but it looked awesome. With future code optimization, I'm sure a single card solution from the 8800 or X2900 series will run it great on fairly high settings(not to mention it was demoed at HDTV resolution 1080i/p). The developers said the hardware support scales back 2 years and forward a year and a half if I recall properly. Crysis was built on DX10 from the start and is the flagship game for the new API, expect all other DX10 games to come with the same range of requirements.

Also, don't panic about performance for now. A developer was quoted saying the game ran very well on his X1900XT and most people say the DX9/10 versions are very similar. :) 
June 6, 2007 5:49:01 AM

Quote:
Oh FFS. You haven't even seen a native DX10 game yet... you're basing your conclusions on some half baked DX10 demos and patches. Also, I have to wonder if you've seen the latest DX10 nvidia demo in action: Human Head

I think you'll find that when Crysis and Alan Wake is released you'll be regreting this post. DX10 is biggest change in API's ever for developers.

I won't argue that M$ attempting to use DX10 to sell Vista, but lumping nVidia and AMD into the "conspiracy" is absurd. Read all the comments every game programmer has to say about DX10 and then come back in here and comment. They all love it.


Still drinking the kool-aid from marketing, eh? Relic was working on the CoH patch since before CoH shipped. Same with Flight Simulator from Microsoft. Pretty impressive Relic managed to get their patch out first IMHO. Lost Planet had the 10 demo out at the same time as the 9 demo. CoH is and always has been d3d10 native.

10 may be a big API change causing major rewrites and delayed, or cancelled, projects, but it offers less than 7 to 8 or 8 to 9. Less than SM2 to SM3 even. Why do you think Supreme Commander dropped 10? 10 is a very small change, your GPU limited game is still going to be GPU limited!

If you're ignoring CoH and LP, all the d3d10 SDK samples, every real fact about d3d10 based on no concrete information about Crysis other than the developers telling you to get really excited and please always talk about their game and spread the hype about their game you're a marketing pawn.

Crysis and Unreal will look virtually identical in 10 and 9 just like CoH and LP.
June 6, 2007 5:54:17 AM

It's obvious you're not a developer...

Changing versions of DirectX isn't exactly an easy proposition... quite a bit of your code ends up having to be rewritten to take advantage of the new version. Hence the importance of waiting to see games that are being designed from the ground up with DX10 in mind, as CoH and FSX were both originally DX9 games, and Lost Planet, as a 360 title, uses a variant of DX9 as its base version. The first native DX10 game isn't going to be until UT3 and Crysis come out almost simultaneously in September, and there's a reason for it; updating code and taking advantage of new technology in 3D graphics is never easy, and, in terms of capabilities, the jump from DX9 to DX10 is actually one of the largest since the jump from DX7 to DX8...
June 6, 2007 6:58:11 AM

Quote:
It's obvious you're not a developer...
Changing versions of DirectX isn't exactly an easy proposition... quite a bit of your code ends up having to be rewritten to take advantage of the new version.


It's obvious you're not a reader because that is exactly what I said.

It's not easy. In fact it is difficult. MS's FS team has talking about d3d10 for years! CoH and FSX were designed with d3d10 in mind. And both took a long time to release. And it doesn't matter.

Yes, the API itself is different and requires major reworkings of old code bases like porting from Glide to d3d to OpenGL, but going from 9 to 10 offers next to nothing. You will be shader bound still. All d3d10 offers is longer shaders, but SM3 already offered longer shaders than any rumoured next generation hardware can render real time. Why make them even longer for rendering??

Quote:
The first native DX10 game isn't going to be until UT3...


Actually UT3 is a port of the d3d9 xbox 360 Gears of War engine. Epic has been console focused for the last few years and there is much much more money to be made licensing technology there. It's basically a mod for GoW with 10 hacked on to the side.

Quote:
jump from DX9 to DX10 is actually one of the largest since the jump from DX7 to DX8...


What did 8 offer over 7? Tell us everything.
And what did 10 offer over 9? If you can look it up you'll see 10 offers nothing to make games better other than lower CPU overhead which doesn't help low end d3d10 cards in the slightest. It helps people with slow old CPUs. Look it up and "prove" to us when you say something it is true.

Supreme command won't go d3d10 because d3d10 offers nothing. It's just a bunch of work. Developers are realizing this now.

Soon the fanboys will admit it too.
June 6, 2007 7:21:21 AM

Actually, Unreal Engine 3 was designed to be able to support DX10, Gears of War just uses a SM3.0 code path because, well, the 360 can't handle anything higher. :)  That being said, the DX10 renderer has been a main reason why the release date for UT3 has slipped into September...

And, if you don't know what DX8 added over DX7, well... it's kinda the whole point here. Something about programmable shaders or something... whatever, I guess. Shaders that, through DX9, only basically got more complex. DX10 not only allows for greater complexity and flow over DX9 shaders, but it also adds support for a completely new type of shader (the geometry shader) that has the greatest potential to provide differences in visual capability over DirectX 9. Since you didn't know about geometry shaders (or just didn't know that they are new to DX10, whatever), I assume you haven't done much looking into what DirectX 10 brings to the table before you started to disparage it. Is it perfect? No, but it is a significant advancement beyond DX9.
a b U Graphics card
June 6, 2007 7:33:51 AM

Titan talks a good game, but all I see is DX9 coming from him. I agree with you. Fanboy? Naw, I just dont believe that all the devs lied to us. I just dont believe that several articles written by intelligent people were lying either. If the geometry shader is just a myth, no "fanboy" created it.
June 6, 2007 2:16:29 PM

Without reading everyone's posts....I'd have to say that until they can offload parts of the code to CPU/GPU as needed to share the workload, and until everything is optimized, and until they start adding extra cores to the GPU and have native quad CPU's and actually make code that can utilize them...DX10 isn't going to show it's REAL potential. We're still evolving into DX10, this is barely the beginning. It'll be three years until we see the real power of DX10 so your preliminary bullcrap doesn't mean squat. It's like sitting on an Apple2 and talking about what Windows 95 and DX 2.0 can do.
June 6, 2007 7:30:30 PM

Yes you summed it up nicely there Johnny and highlighted the real problem the hardware just isn't there, from a mainstream point of view. Sorry if i came across as if i was dumping on the DX10 API, on the contrary i think its a great idea :)  But it has been pushed rather early i believe. There is unlikely to be serious midrange card that will be able to run DX10 better than DX9 (better = playable fps with better eye candy)
June 6, 2007 7:34:52 PM

I don't think Crysis is DX10 "from the ground up."

Am I to believe the created two seperate projects, a dx9 and a dx10 version so they could have to deal with two seperate QA and bug fix projects? No way.

More likely than not Crysis dx10 will look insignificantly better than dx9 but at a significant performance penalty.

It's hard to see the difference between Crysis dx9 and dx10.

DX10 is all hype. Why?

1. We won't see "true" DX10 for another 2 or 3 years when the majority have dx10 hardware.
2. The visual enhancements are subtle (nicer shadows, better motion blur, volumetric smoke) but the performance cost is high on today's hardware.

I'm pissed because I fell for all this hype. I just couldn't wait for Vista and dx10 and all those fake a$$ dx10 game screenshots that never materialized.

M$ and Nvidia/ATI made fools of us.
June 6, 2007 7:47:28 PM

DX10 is hyped because in order for games to actually be developed and sold it has to have potential buyers. The problem is that people don't want to buy the hardware unless there are applications for the hardware. It's like buying a Sony PS3 without games...there WILL be awesome games that utilize the cell processor, but to get there Sony has to take a huge huge billions of dollars risk and make sure the hardware will last 10 years. Nobody wants to buy a gaming system that doesn't have good games, software developers don't want to write code for hardware they're not sure people will buy, and if the whole process fails Sony is left holding the bill and a lesson hard learned. Stocks plummet upon early release and Sony looks bad. The whole computer market is iffy and tough to maneuver, which is why I'm peeved at the ridicule that goes on of companies like Sony and AMD when they make investments and are unable to compete. Sorry if I seemed a bit brash towards the original post, but how you swing an online forum post can ultimately sway minds of people and alter the future of tech. It used to be that any cool invention was a good one, but not anymore. We're far too demanding of consumers, with too much time and too much cash...or rather, too little in savings. When it comes down to it, DX10 hasn't perverted anything...it's a technological marvel compared to the technologies of the past and it's still in development. Making the transition to it is tough for everyone, M$, consumers, developers, and hardware companies. Give it time, and let everyone work together to bring about this new technological advancement...invest your cash wisely and let those with money to blow forge the way for the rest of us. Someday we'll have gadgets that are 10 times as powerful in the palm of our hands that can create holograms for us to masturbate and virtually beat up. Just give it time.
June 6, 2007 9:20:34 PM

DX10 is an API, an abstraction of hardware for use by application programmers nothing more. So without the necessary hardware its irrelevant. Hardware and software in away both spur each other, hardware is indeed useless without software and conversely there's little point in creating code that is unrunable.

The real headache is with the developers who have to deal with 2 completely incompatiable API versions. There needs to be 2 completely seperate rendering code paths, and this is where the problem lies until there is a significant install base of hardware dx10 will not be cost effective to implement for most devs.

I'm just abit down on the fact its going to take awhile :)  so many things i want to see in games but its going to take time :) 
June 7, 2007 3:27:18 AM

I also have to point out, that your not going to find games that Utilize dx10 well yet, because the cards that the devs needed to test the code for the effect, have only come out recently and the development cycle for most good/great games are at least two years, longer for those done really well or with a large story line or detail. Plus with most games the devs need hardware that can show them all of what they want the game to do with DX at the same time so they can make sure nothing will cause another effect to do something wrong, so again it will take some time just by hardware before you see games that really utilize the eye candy that DX10 can deliver.
June 9, 2007 5:16:07 AM

Quote:
And, if you don't know what DX8 added over DX7, well... it's kinda the whole point here. Something about programmable shaders or something... whatever, I guess.


That was my whole point. You don't understand the significance of what was added. You can correctly memorize words, but you don't know what they mean. That's why I said tell us everything. All you can remember is something about shaders and to try to hide your lack of understanding what the shaders were capable talk like it is so obvious it isn't worth your time.

Explain the top three features of Direct3D 8. Feel free to use the internet figure out what the first two were used for and how things were done in the past.

Quote:
Shaders that, through DX9, only basically got more complex. DX10 not only allows for greater complexity and flow over DX9 shaders, but it also adds support for a completely new type of shader (the geometry shader) that has the greatest potential to provide differences in visual capability over DirectX 9.


Even though you don't understand shaders it blows my mind that people will incorrectly boil things down to something simple and throw logic away. You consider adding one shader type, the GS in 10 over 9 bigger than adding the PS and GS in 8 over 7? And yet this shader type wasn't useful enough to be added right away in SM1. Nor in SM2. And not even SM3. Yet somehow it's going to be really awesome. We just need to wait and see.

It just goes to show you aren't able to accurately respond to my 7 to 8 question.
!