• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

PPU useful NOW?

Is a dedicated PPU useful right NOW?


  • Total voters
    172
  • Poll closed .
Your to early with your judgment. Before you can say it flops or is a sucses you must be able to look back to the full attemp of the try.
Like a 100m sprint. It doesn't matter the first 2 meters. Till the finish line. wich includes a lot of those promised 100+ games being released. With a milestone UT2007 would be.
The more meters the beter indication you get.

Down the line it price and polulairity can change drastacly or dramaticly. just wait and see.
Price could be deversed by in time a more complete produkt line

$300 now. and one game. The first smal step. In a large journey.

Like they compare with 3dFX voodoo with games that are released 4 years after the first Voodoo came out. As if Graw the is only one. For the next 5 years.
 
considering ut2k7 is just effect physics it can't be considered a milestone for Ageia.
 
Yah so? I bought PPU for Graw effecs features. A extra GPU wouldn't do because the Havok engine is only the CPU Software one. You need a PPU for it.

If UT2007 have some done well eye candy Physicsl. You need a PPU for it not a GPU. It will certenly do good for PPU. even without the gameplay Physics.

It would be differen if a large part of the eye candy games uses HavokFX instead a decent amount of games with only PhysX API games. It stands now 100+ vs non for Physics GPU API support.

It's certenly a big plus PhysX do gameplay Physics to. But it boils down by the choice of wich API where the dev's with a killer app are the most important once.
 
ScYcS said:
(<---- always wondered why people sign their post....it's not that we can't spot your name on the post header anyways)

I think it looks groovy :)
 
The way I see it, based on this pole, Ageia isnt as screwed as some say it is. Over 50% think it may be worth $300 at some point. Another 25% would buy it if the price dropped...

compare that to any $300 GPU... I'm not talking full product family. I'm talking like an X1900 or 7950.

edit: I included those that would buy it now, with the ones that think it could be worth it in the future
 
nhusby said:
The way I see it, based on this pole, Ageia isnt as screwed as some say it is. Over 50% think it may be worth $300 at some point. Another 25% would buy it if the price dropped...

You're making a big error in your evaluation.

The sample group you're using is not representative of the average gamer. You're on a hardware enthusiast site. You're dealing with a group of people who spend their time talking themselves out of pointless upgrades, not talking themselves into one.
 
eastvillager said:
You're making a big error in your evaluation.

The sample group you're using is not representative of the average gamer. You're on a hardware enthusiast site. You're dealing with a group of people who spend their time talking themselves out of pointless upgrades, not talking themselves into one.


Hey my upgrade to a 7900gtx is totally needed. I bought my 7600gt when I was running a 17in flat panel monitor. I am now riding a 20.1in widescreen monitor my poor 7600gt can't handle such things so it's not pointless :)

I am sure there are hundreds of thousands of these cards flying off the shelfs to.
 
Include an option for "It could be useful in the future and is not worth $300 now or later" and I'd take that. Further, I don't have room for a PhysX, even if it does make itself useful...K8N Diamond Plus is stuffed to the gills with 2x7900GTX and an X-Fi. I need a PCI slot liberator, like the powerstrip extenders except for a PCI slot.
 
not usefull for me now. i can see later when we will need them, but the price will need to come down signifigantly. i know its like beating a dead horse, but the the only thing this card can do is make better physics. everything thing else in the computer helps withg other things. for the price i could buy an x-fi, another gig of ram, and ,maybe a better mouse. plus i cant fit one in my mobo right now. i ahve my wifi card, and my maze4 gpu cooler. the maze 4 cooler blocks all the of the pci slot below it, and the middle pci slot is partially blocked and will ony exept small pci cards
 
Not now, not in the future, not for $100. Any way it is cut, it could be done on the software end. All it takes to do it is math and variables for things like weight/dimensions/resistance. Regardless, it's all an attempt at combining eye candy with realism. If I really wanted that as my top priority, I'd just buy movies. Damn, that would save me a lot of money, too! <shakes fist at computers>
 
eastvillager said:
You're making a big error in your evaluation.

The sample group you're using is not representative of the average gamer. You're on a hardware enthusiast site. You're dealing with a group of people who spend their time talking themselves out of pointless upgrades, not talking themselves into one.

I should have been more specific in my point.

my point is that of the enthusiast market, 75% are open to buying at PPU.
So if a hardware company can effectively make a profit on a chip that only sells to under half of the enthusiest market then Ageia shouldnt have much problem...
 
Well all new thing start with the Hardcore gamers and hardware enthausiast. The masses don't know about the card yet. They no talk about it on general oriented PC related site.
It must grow. Voodoo 1 at release was not made for the poor student. It was made for the Hardcore gamer who is prepered to pay for the next revolution for games. Here is where a new market started the 3D game accelerators. Wich after lot of years they comes now in onboard solution to cheap to extreem ultra dual variants of competion that wo the fight.

It's new and and as always the Software lacks behind. Knowing that a average Game take around 1 to 5 years to get done. We only can hope if it becomes a sucses after at least 2 years.

8 people liked it. including me, we are the early adopers.

mrjminer said:
Not now, not in the future, not for $100. Any way it is cut, it could be done on the software end. All it takes to do it is math and variables for things like weight/dimensions/resistance. Regardless, it's all an attempt at combining eye candy with realism. If I really wanted that as my top priority, I'd just buy movies. Damn, that would save me a lot of money, too! <shakes fist at computers>

Wat can be done?
You can cranck up Physics a 1000 times. compared to what is done now in software.
If some Highe Profile Developer put that lots of Physics hardware accelretad Physics to very good use that makes its game stands out. Other Dev's for competition can't stay behind.

It means. instead waitin how Physics computation in software grows with the increase of power of new CPU. Instead it get a large Hardware boost to Put Physx where CPU would be 5 to 7 years later.
And till then a PPU would be then even more powerfull like a P3 or something from the competition. If they take a 2 year produkt cycle.

If ageia made to the years. Physics get a large step. Wich is on a role because ATI nV MS Ageia making hard work of it. So it coming.
 
Here's what I want to see: PCIe card with a CPU socket on it. There, physics solved.
 
No I think not. For iNTel it means a Northbridge on it to. Wich uses PCI-E for Intervhip connection.

AMD with HT interchip standard has a solution for it for heavy computing with a array of co PU's With AMD sockets The CoPU gets its own HT link and dedicated Memory Though a Other CPU Socket on the Mobo. It can be used by a Opteron Or this CoPU
The register Opteron Copro
DRC Ccomputer

They could do this with Ageia also Onley a Dual CPU mobo needed.
 
no option for useful later not worth 300 later?

whats the direction are they going with this? is this supposed to be mainstream or ultra high end? is this supposed to compete with SLI/CF?

Im a midrange guy who buys the $350 GT single card. At this price, i dont see me buying one at all. If this is supposed to compete with SLI/CF and i only had enough resources for one upgrade, id choose SLI/CF hands down even tho im not even remotely in that market? why? because as i competetive gamer, i value smooth framerates/good IQ over realism. What does a ppu give me over the competetion? Now if something like cell factor or graw was in CAL and i couldnt compete because of lag/no ppu, i dunno where i stand.

an example would be if i were competing in a driving game ladder for times. Would i use a force feedback steering wheel? hell no. Why? because its easier to be precise when ur steering wheel isnt shaking uncontrollably trying to provide a realistic representation of driving. Now if i can turn off fog/shadows/extras explosions to help get rid of distractions in FPS's, id do it (legally of course). Its my understanding that PPUs just add that kinda stuff which is counter productive to my goals. However if this were impletemented in a showcase single player game like oblivion, id see no prob with that. But then again, single player games have low replay value to me and i hardly buy them anyways now. Just a quick leech for a few days till i beat it then its off my rig. Am i really gonna pay 300 bucks to support games i play infrequently/dont buy and hinders games that i play alot and buy like online FPS for competetion and MMOS?
 
Oh yes PPU could be worth $300 later. But offcourse not for everyone.
Because there is more then the Fast action FPS online genre.
There are also slow peace tactical shooter where you are more aware of the scenery.
Also RTS with lot of realistic behaving unit's
Race games with realistic behavior more aimed at simulation. Not every one likes it but a not so small part preferes simulation instead of simple arcade games style of race agains you friends.

Different people different needs and budged and play different games.

I dont play much online. Online games I play often with Bot's, COD I play Singleplayer only. Only game I played online was AA:SF but not so regulary.

Where a Frag fest game gives 90fps you the edge of sucses. With the cost of low IQ for FPS.
In a tactical the FPS can drop to 15fps and still be playable.
 
“Now if i can turn off fog/shadows/extras explosions to help get rid of distractions in FPS's, id do it (legally of course). Its my understanding that PPUs just add that kinda stuff which is counter productive to my goals.”
In BetOn Solider the PPU appears to be speeding up all physics offloading from the CPU and boosting FPS. If you turn off the extra effects you should end up faster with the PPU then without.
 
Pottsey said:
“Now if i can turn off fog/shadows/extras explosions to help get rid of distractions in FPS's, id do it (legally of course). Its my understanding that PPUs just add that kinda stuff which is counter productive to my goals.”
In BetOn Solider the PPU appears to be speeding up all physics offloading from the CPU and boosting FPS. If you turn off the extra effects you should end up faster with the PPU then without.


so your saying with a ppu, if i turn down eye candy but leave the ppu to do its job, i can have smoother gameplay and more fps? Id pay $100 along with my $300gt card for it if it was this situation. My clan is playing the next big game in CAL and it uses ppu heavily. Under my $300gt class card during that time, i cannot produce acceptable frame rates (60+) during the most intense moments at my monitors native rez with mild eye candy that helps me distiguish things such as enemies,projectiles,special power ups etc.. Adding the ppu will bring me up to satisfactory levels when, going sli or buying at gtx for $150 more wont.

Right now i play CAL for BF2, my 6800gt will give me 60-80fps at 12x10 native/2xaa/4xaf on my 17in monitor. BF2 doesnt support ppu. Do i need a $300 ppu? hell no, but will see in the future. It all depends on price/performance as well as necessity. All of those questions need to answered especially on heavily anticipated games like quake wars and ut07.
 
Also notice that once was 16bit color depth the standard and 32bit luxuray. Now is 32bit the standard.
You play with a old SM3.0 6800GT

Hardware evolves new eyecandy and effects features are introduced. But to use it to its full you need the latest hardware. And that would be a modern High-end card. Like a X1900XTX 512MB. A next gen GPU can often accelerate once new features way better and a GPU refresh put another speed bump.

Just like FSAA give in first days a heavty Performance hit. this hit is still reduced. FSAA is becomming Mainstream. same goes for AF

One day Hardware Physics will be mainstream Ageia or not and for a large target platform dev's must need to put it to good use because Dev's and there title compete also with each other.

It means Physis is standard and on day will be a requierment. And of course a full range produck line. also for the budged gamers with just a 17" monitor
 
“so your saying with a ppu, if i turn down eye candy but leave the ppu to do its job, i can have smoother gameplay and more fps? Id pay $100 along with my $300gt card for it if it was this situation.”
Yes but only with the games that use the PPU to do all the physics that the CPU used to do. So if you turn off the extra PPU only physics then do all the CPU physics on the PPU you free up the CPU too boost FPS when CPU limited. But not for every game, it seems to work in BetOn solider but not Switchball.
 
I don't understand why Ageia hasn't released any software to show off what the PhysX cards can do... I mean, I thought that NovodeX Rocket demo was largely a plug for their cards, since that exploding tower demo with ~15,000 bricks was beyond the capability of any processor at the time (maybe still is... I'm still on an A64 3000+ (on which it runs at ~4fps), so I couldn't tell you...).

It'd also give us some idea of what the card is actually capable of, since it's very easy to create your own demos, of arbitrary scale, with .lua scripts. And all the predefined objects are geometrically simple and untextured by default, so GPU limitations shoudn't be an issue - it'd make a perfect benchmark for the true capability of the PhysX card.

Maybe that's what they're afraid of... They haven't really shown us anything so far to support some of their claims (namely 30,000 physically simulated objects), and maybe there's a good reason...
 
LuminaryJanitor said:
I don't understand why Ageia hasn't released any software to show off what the PhysX cards can do... I mean, I thought that NovodeX Rocket demo was largely a plug for their cards, since that exploding tower demo with ~15,000 bricks was beyond the capability of any processor at the time (maybe still is... I'm still on an A64 3000+ (on which it runs at ~4fps), so I couldn't tell you...).

It'd also give us some idea of what the card is actually capable of, since it's very easy to create your own demos, of arbitrary scale, with .lua scripts. And all the predefined objects are geometrically simple and untextured by default, so GPU limitations shoudn't be an issue - it'd make a perfect benchmark for the true capability of the PhysX card.

Maybe that's what they're afraid of... They haven't really shown us anything so far to support some of their claims (namely 30,000 physically simulated objects), and maybe there's a good reason...

Try this:
Cubewall

"This is a demo of ~3k cubes stacked as a wall. WASD moves camera, left mouse button turns camera, SPACE shoots a ball and P pauses the physics simulation (useful for seeing how much physics impacts performance). Software and hardware physics supported. DirectX 8.0 compliant GFX card required"

Terra - It can run with...and without PhysX hardware...but what a difference ;)
 
LuminaryJanitor said:
I don't understand why Ageia hasn't released any software to show off what the PhysX cards can do... I mean, I thought that NovodeX Rocket demo was largely a plug for their cards, since that exploding tower demo with ~15,000 bricks was beyond the capability of any processor at the time (maybe still is... I'm still on an A64 3000+ (on which it runs at ~4fps), so I couldn't tell you...).

It'd also give us some idea of what the card is actually capable of, since it's very easy to create your own demos, of arbitrary scale, with .lua scripts. And all the predefined objects are geometrically simple and untextured by default, so GPU limitations shoudn't be an issue - it'd make a perfect benchmark for the true capability of the PhysX card.

Maybe that's what they're afraid of... They haven't really shown us anything so far to support some of their claims (namely 30,000 physically simulated objects), and maybe there's a good reason...
Like the need for fun gameplay with 30.000 object.
I think it would be some amount but with a lot of diverent Physics effect simultainlously.

Cellfactor is lot of Boxes but Cloths and fluid to. Put a babe in with long hair to her waist.:) Just another effect.
and more you push the PPU to it limits.

30.000 object in one Physics dissipline is nice for 3Dmark junkies.

ATI has performance claims but with what mix of physics and what kind of load?

PPU could be good on a some Physics like boxes and GPU more on hair. as example.

Physics benching. Would be difficult. game could take any mix they like plus each of them could be disable or anabled.
Like long hair babe or a Bold headed babe.:)
Plus if you want to bench reallife games. They must support both solution. Like Games Where you could choose for Glide/DX/OGL
 
Back
Top