physx in ut2k7

[sas*w)v]

i remember seeing a while back that ut2k7 was one of the titles that was going to implement some physx acceleration. Anyone know if this is still the case and what/how much will be used? I'm holding off buying one of these cards untill a killer game is released for it...and 2k7 could be just that title if it actually is enhanced with the use of a physx card, unlike a few of the other games that are around atm. Buying the game anyway tho

[AFG34]

yas, it will make use of physix card.

[Piglet]

Hopefully it won't be one of the games where the physix card actually LOWERS framerate.

[Nny]

Amg Piglet How Is Life In The Uk!?!

[Hedge-o-Matic]

Hopefully it won't be one of the games where the physix card actually LOWERS framerate.

Like the Soundblaster Live support? Or the EAX support? I agree. Too often, support for such specialized technologies makes the game nigh unplayable, and have to be disabled. Can someone please explain why hardware that has its own chipset designed for specialized tasks slows down the game? WTF? Is this more a matter of crappy front-end bus speed, or lousy integration by Epic? I'd wait before investing in Physix, that's for sure.

[MonsOlympus]

hmmz, on the alienware site there newest pimp rig doesnt let you put in a creative card and a physx. Mind you theres quad sli in there too but you'll need it if there was a fps hit from having these

[iLL]

I too was thinking of a phys card for ut2k7,but I will wait until I hear some feedback on how it impacts the game both in terms of performance and visual payoff.

Too many other things to upgrade first.A physics card will be last on my list.

[_Lynx]

What Mark rein said that the game will run even w/out PhysX card, as PhysX engine works nice even in softwae mode, but they'll look into making few maps that will require PhysX to run.

[sphinx]

haha w)v!! lost your virginity inhere aye?

say hi to djero :<

[Phopojijo]

Like the Soundblaster Live support? Or the EAX support? I agree. Too often, support for such specialized technologies makes the game nigh unplayable, and have to be disabled. Can someone please explain why hardware that has its own chipset designed for specialized tasks slows down the game? WTF? Is this more a matter of crappy front-end bus speed, or lousy integration by Epic? I'd wait before investing in Physix, that's for sure.

Soundblaster Live support? Yea, thats weird. EAX support? How is that unexpected? You're processing in realtime sounds based on geometric properties defined by the drivers (echo range, etc.) -- of course you're going to take up extra resources.

As for PhysX... the performance hit is NOT accurate... there's a performance enhancement when run in software mode versus Hardware mode.

If you get 70FPS in Software mode on UT99, and you install a GeForce2 then crank it up to 1600x1200 everything up... do you expect the framerate to be higher than 70? Of course not...

Sure you're offloading the burdon onto a separate device... however, that separate device doesn't take EVERY burdon away from your videocard and CPU... just some of it. Think of it this way... every particle that the PhysX card deforms is like a vertex-animated model being rendered like a static mesh. More performance? Indeed... but your videocard will STILL need to RENDER the end result... which is more detail than the former.

[Aggressor]

I tested PhysX card in GRAW. Yes, it caused a clearly observable hit on performance. The reason seems to be the fact that higher physics settings (tailored for hardware acceleration) also meant more geometry and effects in view and therefore you're actually causing the game to temporarily have far greater detail in comparison to lower physics settings.

So, if you by any chance have PhysX card (I myself warn against buying it yet), make sure you also have top-of-the-line gear in all other departments.

[sas*w)v]

haha w)v!! lost your virginity inhere aye?

say hi to djero :<

although from reading some posts/replies on this message board it appears that a hell of a lot of the people here actually have no idea what makes a good ut game and i hope that epic ignore most of it. Far more relevant stuff is posted on unrealnorth or esreality tbh

[sphinx]

yeah, but keep in mind that the unrealnorth and esreality crowd are on average quite much more evolved into the scene than the avg forummer here :/ sadly

[Hedge-o-Matic]

Gee, I guess we on this forum should just keep our ignorant mouths shut, then, since we're not as "evolved" a "forummer" [sic], and since our posts here are, as you say so elequently, "not relevant".

I can see how talking about the physx capabilities of 2k7 would be considered by the elite, "evolved forummer" to be seen as... oh, I don't know... not lofty enough.

But, given that this forum is, much like the others you mentioned, running on rumor and specualtion at the moment, this primitive, almost stone-age topic will have to do, until we can play the game, and perhaps join the upright-walking nearly-mythical supermen on those other forums. Until we somehow -like the ape-like primates in 2001: A Space Oddessy- are given "UT-sentience", we can only hope our very dim hopes, and write our faint, indistinct and half-formed ideas to one another. In any case, our dim, unenlightened thoughts are irrelevant, since we, as you say, don't actually have any idea what make a good UT game. When you say that Epic should ignore us, you are so right! Clearly, we'd be better off just listening to other, wiser voices than our own.

Thank you for joining the community here, and making it all so clear for us.

[iLL]

although from reading some posts/replies on this message board it appears that a hell of a lot of the people here actually have no idea what makes a good ut game and i hope that epic ignore most of it

What makes a game good,great, or suck is in the eye of the beholder and is ALL opinion based.

Just as a persons view onto another being a friendly,easy going,open minded gent OR a big headed,think you're better than everyone else ass jockey.

Far more relevant stuff is posted on unrealnorth or esreality tbh

Goodbye then.Thanks for droppin in.Don't forget to tell your friends.

[iLL]

yeah, but keep in mind that the unrealnorth and esreality crowd are on average quite much more evolved into the scene than the avg forummer here :/ sadly

Thats the competitive community and Epic has already gave the vibe that they intend to appeal to the masses.

I am in no way knocking those sites because they are indeed filled with greatly skilled and good hearted individuals.Just so happens Sas*w does not seem to be one of those,or maybe he is just in a bad mood.

[GiantWeirdo]

Hmm, just read up on this Physx card.

Interesting idea, but it won't last long. My guess is it will go the way of the math co-processor. Some of the older members might remember the days of the 386 and before where the math coprocessor could be added on for better floating point performance. Then the SX and DX variants of the intel processors came...
This physx card might offer a boost, but I would think making the games multi-threaded would be a better way to improve physics. Using the multi-core processors that people have now should allow for some substantial improvements in game physics. Maybe not as good as the physx card for now, but soon, as processor power increases, people will wonder why they need to take up a PCIe slot for this device...

[The_Deacon]

Hmm, just read up on this Physx card.

Interesting idea, but it won't last long. My guess is it will go the way of the math co-processor. Some of the older members might remember the days of the 386 and before where the math coprocessor could be added on for better floating point performance. Then the SX and DX variants of the intel processors came...
This physx card might offer a boost, but I would think making the games multi-threaded would be a better way to improve physics. Using the multi-core processors that people have now should allow for some substantial improvements in game physics. Maybe not as good as the physx card for now, but soon, as processor power increases, people will wonder why they need to take up a PCIe slot for this device...

Well the same could be said for graphics cards really, but they are currently still going strong. PhysX cards have hardware optimised for physics calculations, like graphics cards have hardware optimised for graphics calculations whereas a processor is a more generic processing unit. Multi-core processors will probably be used in the way you describe, but I doubt that it will match a dedicated physics card, at least not for a while.

[Hedge-o-Matic]

I can understand the idea behind dedicated graphics hardware benfiting from particular chipsets, since graphic capabilites are something of a known element. There's the idea of polygons, textures, lighting, and so forth, and the math involved in getting this on-screen is understood becasue it's going to be the same all the time. Regardless of what sort of image you're showing, you're still dealing with the same sorts of 3D data.

Physics simulation, however, can be arbitrarily complex. I can't seem to understand what standard mathematical processes can be put onto a chip for dedicated acceleration, especially since realtime physics simulation is a pretty new field.

Are physix cards just optimized for mass amounts of velocity / vector calcuations? Can they really help in general solutions (for example, in rendered 3D scenes, such as 3D MAX or Maya, or specialized simulations), or are they geared specifically to the somewhat quick-and-dirty world of 3D gaming?

I'd be interested to know more about them, and what mathematical processes they optimize, rather than just hype sheets and press releases on how awesome the idea is.

[Hanji]

Hmm, just read up on this Physx card.

Interesting idea, but it won't last long. My guess is it will go the way of the math co-processor. Some of the older members might remember the days of the 386 and before where the math coprocessor could be added on for better floating point performance. Then the SX and DX variants of the intel processors came...
This physx card might offer a boost, but I would think making the games multi-threaded would be a better way to improve physics. Using the multi-core processors that people have now should allow for some substantial improvements in game physics. Maybe not as good as the physx card for now, but soon, as processor power increases, people will wonder why they need to take up a PCIe slot for this device...

The game engine IS multi-threaded.

[Locust Bait]

I wouldn't be judging this too quickly guys, have you ever seen the chip used on the UE3? Lets just wait, thats what I'm doing. Is the chip worth it? If not, I'm getting a better sound card.

[GiantWeirdo]

The game engine IS multi-threaded.

Of course as all new games should be. I was just saying multi-threaded coding is where they should focus their efforts to improve physics modelling.

My opinion is just that Fast CPU today + Physx card = Fast CPU 6 months from now with no physx card. These physx cards will be obsolete so quickly and you'll have to keep upgrading to the physx2 and physx3 just to keep pace with what could have been done with just a CPU in the first place, just slightly slower at the time.

Its an interesting idea, and I could be wrong, but as I pointed out, the CPU world has been down this path before with math co-processors. If there really is anything special about the physx card, those same commands and hardware functionality will be added to the next versions of SSE and 3DNow!.

[Boksha]

I wouldn't be judging this too quickly guys, have you ever seen the chip used on the UE3?

It's pretty much 100% sure the engine (and thus the game) will support the PPU.
The real question is, will UT2007 have enough physics-driven objects to notice a difference if the physics are done by an addon card? For now, I wouldn't count on it. UT2007 is a multiplayer game, and complicated/chaotic physics and internet do not mix very well, so there's a pretty good chance most of the physics driven objects will just be details you can turn off in the menu.

[sas*w)v]

What makes a game good,great, or suck is in the eye of the beholder and is ALL opinion based.

Just as a persons view onto another being a friendly,easy going,open minded gent OR a big headed,think you're better than everyone else ass jockey.



Goodbye then.Thanks for droppin in.Don't forget to tell your friends.

Ok dont get too upset if you think my statement applied to you (maybe it does, maybe it doesnt - i dont know). The point is that a lot of the people's suggestions here really arent thinkin about their actual ideas in the context of a mulitplayer game and how it can affect things.

It's a shame that more people from esreality/unrealnorth dont post here as it could be educational to a lot of people here who, it appears, mostly play this game in single player or at a ffa level online - which is fine if thats all they want to do, but if some of their ideas are taken onboard it can damage the high level competitive end of the game.

[Jezza101]

It's a shame that more people from esreality/unrealnorth dont post here as it could be educational to a lot of people here who, it appears, mostly play this game in single player or at a ffa level online - which is fine if thats all they want to do, but if some of their ideas are taken onboard it can damage the high level competitive end of the game.

But conversely we dont want the game to be designed purely towards competitive gamers - lets keep the fun in there too! I think this forum actually has a good cross section of players so is therefore an ideal forum for discussion.

[Nietzsche]

But conversely we dont want the game to be designed purely towards competitive gamers - lets keep the fun in there too! I think this forum actually has a good cross section of players so is therefore an ideal forum for discussion.

Why do people always think, that competitive games are no fun? A game that works for competitive-gaming is balanced, therewith in general more fun than games that are not playable competitively.

[Boksha]

Why do people always think, that competitive games are no fun?

It's very possible to design a game that's good for competetive play but not really much fun for casual players and vice versa. (I'm thinking Warsow as an example of the former, and any WW2 shooter as an example of the latter, although in all honesty you CAN play those competetively; I just don't think it's much fun because it tends to be all-camp-no-action)
Of course, a great game is fun for all, although it doesn't have to be fun if it's played between badly matched opponents.