Get your PhysX card now...

[CP5670]

Benchmarks arent the issue, the card can clearly do things which arent possible without dedicated hardware. Its not "pretty stuff" at all, as its something that genuinly affects the way the game works, its not about graphics.

This is the main problem I see with the physics card idea. I mean, with a low end or integrated video card, you may have to play with a low resolution and choppy framerate, but the gameplay itself is still exactly the same as what someone with the baddest quad SLI setup would be getting. But if this physics card alters the game mechanics directly, the people who have these cards and those who don't have them would really be playing quite different games. This would especially become a problem in multiplayer games, where things need to stay synchronized for everyone, so either the physics card would be a requirement for everyone (and the game would sell like crap) or it would be reduced to just doing graphical effects (like swaying grass or something, stuff that looks nice but doesn't affect gameplay at all).

I like the idea of having a dedicated physics card, but I don't get how it's going to work in practice unless games start requiring it, which seems unlikely for quite some time. I can actually see a lot more uses for this in numerical computation apps than in games and it may be worth it to me just for that (although I definitely wouldn't spend $250 for it).

 

[Genx87]

Originally posted by: munky

Originally posted by: concernedsophist
I am so surprised at how many ?anti-physics people? are around here! Come on people. Obviously it will enhance game play! Even if the first generation of cards is overpriced and underutilized it is a stepping stone to bigger and better things (read cheaper and more powerful). So don?t buy the $300 card this year. However, it is a necessary step to make these cards to make advances in physics hardware, software, and market penetration. Imagine the future games where the cpu is almost entirely free to do AI! This will lead to huge advances in game play. How can you be a computer and game enthusiast and not be excited about the prospect of this technology, even if you are not going to purchase an early version of the hardware?

I'm an enthusiast, but I'm not some rich fat bastard with more money than brain cells. Ageia is in this business only for one reason - to make money. They tell you that a physx processor will make games so much more realistic because it's in their best interest. No games even fully use the two cores on my cpu, and you want me to believe that the cpu is too weak and slow to handle physics? What kind of physics does a game require to warrant a dedicated processor? Until I see a game that justifies the cost of a physx chip, I will not support it.

CPU's are general pupose processors. When was the last time you saw a general purpose processor outperform a dedicated processor?

Of course they are in the business to make money but what does that have to do with the technology behind it? Clearly games with more physics will be > than games that are void of it.

 

[Genx87]

I like the idea of having a dedicated physics card, but I don't get how it's going to work in practice unless games start requiring it, which seems unlikely for quite some time. I can actually see a lot more uses for this in numerical computation apps than in games and it may be worth it to me just for that (although I definitely wouldn't spend $250 for it).

Chances are devs will scale the engine to the capability of the user like they do with their rendering paths. What a full DX9 card renders and a DX8 or DX7 card renders can be very different. Similarily if game devs start making walls implode, the people who have a PPU will see the wall implode in detail where people who cant handle it may just see a wall have a small animation and turn into rubble.

 

[CP5670]

That's just what I meant though when I said that it would be doing purely graphical effects. Either the explosion/rubble wouldn't actually affect the game in any way (it would be a cool looking effect, not unlike the AA that video cards do), or it would make the gameplay mechanics different for everyone.

 

[Creig]

That's my thought as well. For now, it will simply enhance graphics. Perhaps in the future, games may require a PPU so that the computed effects will be interactive.

 

[Acanthus]

Originally posted by: Genx87

Originally posted by: Sudheer Anne
people don't want to buy more add-in cards, and for those with SLI setups it's likely that all their slots are unusable anyways. if this card was under $100 people might jump on it, but $299+ is a pricetag almost nobody will pay. if they keep prices where they are they are simply shooting themselves in the foot.

It really depends on the titles they have lined up. If they have compelling titles people will buy it, including myself.

My only concern however is not having enough PCIE slots but I am thinking we should start seeing more that two x16 slots on these MBs. If not this is a great opportunity for Nvidia to get their duo cards out for people who want to SLI, but also want a PPU.

3 big MMOs will support it. The hardcore players in those will almost certainly pick up the card. Im taking a serious look at warhammer because i loved daoc for its first few years before mythic destroyed it.

Shadowbane II (will probably bomb like the 1st), Warhammer Online, and Vanguard: Saga of Heroes.

 

[Topweasel]

Originally posted by: munky

Originally posted by: concernedsophist
I am so surprised at how many ?anti-physics people? are around here! Come on people. Obviously it will enhance game play! Even if the first generation of cards is overpriced and underutilized it is a stepping stone to bigger and better things (read cheaper and more powerful). So don?t buy the $300 card this year. However, it is a necessary step to make these cards to make advances in physics hardware, software, and market penetration. Imagine the future games where the cpu is almost entirely free to do AI! This will lead to huge advances in game play. How can you be a computer and game enthusiast and not be excited about the prospect of this technology, even if you are not going to purchase an early version of the hardware?

I'm an enthusiast, but I'm not some rich fat bastard with more money than brain cells. Ageia is in this business only for one reason - to make money. They tell you that a physx processor will make games so much more realistic because it's in their best interest. No games even fully use the two cores on my cpu, and you want me to believe that the cpu is too weak and slow to handle physics? What kind of physics does a game require to warrant a dedicated processor? Until I see a game that justifies the cost of a physx chip, I will not support it.

People think that graphics take alot to render but the fact is true to life physics requires more proccessing then that. Something as simple as wind blowing leaves around would bring a proccessor to its knees These are thing that have to be scripted. AMD has made two Super computers for two F1 teams just to handle air flow over a car. With as many points in a game that need to be computed and each objects reaction with other computed objects a single (or dual) CPU just can't handle it, because the are designed to handle one object at a time really fast. They are made to be general purpose and the speeds they are going they are but if you look at a GPU (which is closer to being a PPU) the 7900GTX is 24 pipes (threads) working at 650MHz for a CPU that would like having a 15 GHz CPU. The only CPU close is the Cell Proc which is why Sony licensed the tech to use those FP cores.

But the basic point here is no matter what device you pick for the task once this tech gets rolling will need to be completely dedicated to Phyics. Which means buying a second CPU, a second GPU, or a dedicated PPU. I think alot of people will be surprised at what this tech could mean and we are less then a year away from having a GLquake of sorts convince the industry that this isn't a gimmick. So munky you play with you Millinea or Virge while we Physics believes will be playing with or Voodoo1's. Do I think It will be as big a deal as that was? No, but I believe it will be close.

 

[AzNPinkTuv]

Originally posted by: Acanthus

Originally posted by: Genx87

Originally posted by: Sudheer Anne
people don't want to buy more add-in cards, and for those with SLI setups it's likely that all their slots are unusable anyways. if this card was under $100 people might jump on it, but $299+ is a pricetag almost nobody will pay. if they keep prices where they are they are simply shooting themselves in the foot.

It really depends on the titles they have lined up. If they have compelling titles people will buy it, including myself.

My only concern however is not having enough PCIE slots but I am thinking we should start seeing more that two x16 slots on these MBs. If not this is a great opportunity for Nvidia to get their duo cards out for people who want to SLI, but also want a PPU.

3 big MMOs will support it. The hardcore players in those will almost certainly pick up the card. Im taking a serious look at warhammer because i loved daoc for its first few years before mythic destroyed it.

Shadowbane II (will probably bomb like the 1st), Warhammer Online, and Vanguard: Saga of Heroes.

was just gonna point that out for the haters saying multiplayer wouldnt support it as well lol

 

[linkgoron]

Originally posted by: Topweasel

Originally posted by: munky

Originally posted by: concernedsophist
I am so surprised at how many ?anti-physics people? are around here! Come on people. Obviously it will enhance game play! Even if the first generation of cards is overpriced and underutilized it is a stepping stone to bigger and better things (read cheaper and more powerful). So don?t buy the $300 card this year. However, it is a necessary step to make these cards to make advances in physics hardware, software, and market penetration. Imagine the future games where the cpu is almost entirely free to do AI! This will lead to huge advances in game play. How can you be a computer and game enthusiast and not be excited about the prospect of this technology, even if you are not going to purchase an early version of the hardware?

I'm an enthusiast, but I'm not some rich fat bastard with more money than brain cells. Ageia is in this business only for one reason - to make money. They tell you that a physx processor will make games so much more realistic because it's in their best interest. No games even fully use the two cores on my cpu, and you want me to believe that the cpu is too weak and slow to handle physics? What kind of physics does a game require to warrant a dedicated processor? Until I see a game that justifies the cost of a physx chip, I will not support it.

People think that graphics take alot to render but the fact is true to life physics requires more proccessing then that. Something as simple as wind blowing leaves around would bring a proccessor to its knees These are thing that have to be scripted. AMD has made two Super computers for two F1 teams just to handle air flow over a car. With as many points in a game that need to be computed and each objects reaction with other computed objects a single (or dual) CPU just can't handle it, because the are designed to handle one object at a time really fast. They are made to be general purpose and the speeds they are going they are but if you look at a GPU (which is closer to being a PPU) the 7900GTX is 24 pipes (threads) working at 650MHz for a CPU that would like having a 15 GHz CPU. The only CPU close is the Cell Proc which is why Sony licensed the tech to use those FP cores.

Errr, no it isnt. a Dual core A64 running at 2.6 isn't the same as a single core A64 running at 5.2
And the cell will die a horrible death (imo).

But the basic point here is no matter what device you pick for the task once this tech gets rolling will need to be completely dedicated to Phyics. Which means buying a second CPU, a second GPU, or a dedicated PPU. I think alot of people will be surprised at what this tech could mean and we are less then a year away from having a GLquake of sorts convince the industry that this isn't a gimmick. So munky you play with you Millinea or Virge while we Physics believes will be playing with or Voodoo1's. Do I think It will be as big a deal as that was? No, but I believe it will be close.

:thumbsup:

 

[Acanthus]

In a specialized processor performance will scale nearly linearly.

with 2000 things being interacted with in a scene you think 16 or more cores cant be fully utilized? Come on now.

 

[flashbacck]

I'm leaning towards the "PhysX card will Flop" category. I'd rather have a multi-core CPU that'll be used in more programs than a dedicated card that'll be used in a few games. But to be fair we need some benches to make a good judgement. Something like:

$250 PhysX + $125 A64 3000+
vs
$355 A64 X2 4200+

(prices from RTPE)

Any games tested would have to be optimized for both the physx card and multicore cpu's, of course.

 

[Energinerbuzzy]

Originally posted by: Topweasel

Originally posted by: munky

Originally posted by: concernedsophist
I am so surprised at how many ?anti-physics people? are around here! Come on people. Obviously it will enhance game play! Even if the first generation of cards is overpriced and underutilized it is a stepping stone to bigger and better things (read cheaper and more powerful). So don?t buy the $300 card this year. However, it is a necessary step to make these cards to make advances in physics hardware, software, and market penetration. Imagine the future games where the cpu is almost entirely free to do AI! This will lead to huge advances in game play. How can you be a computer and game enthusiast and not be excited about the prospect of this technology, even if you are not going to purchase an early version of the hardware?

I'm an enthusiast, but I'm not some rich fat bastard with more money than brain cells. Ageia is in this business only for one reason - to make money. They tell you that a physx processor will make games so much more realistic because it's in their best interest. No games even fully use the two cores on my cpu, and you want me to believe that the cpu is too weak and slow to handle physics? What kind of physics does a game require to warrant a dedicated processor? Until I see a game that justifies the cost of a physx chip, I will not support it.

People think that graphics take alot to render but the fact is true to life physics requires more proccessing then that. Something as simple as wind blowing leaves around would bring a proccessor to its knees These are thing that have to be scripted. AMD has made two Super computers for two F1 teams just to handle air flow over a car. With as many points in a game that need to be computed and each objects reaction with other computed objects a single (or dual) CPU just can't handle it, because the are designed to handle one object at a time really fast. They are made to be general purpose and the speeds they are going they are but if you look at a GPU (which is closer to being a PPU) the 7900GTX is 24 pipes (threads) working at 650MHz for a CPU that would like having a 15 GHz CPU. The only CPU close is the Cell Proc which is why Sony licensed the tech to use those FP cores.

But the basic point here is no matter what device you pick for the task once this tech gets rolling will need to be completely dedicated to Phyics. Which means buying a second CPU, a second GPU, or a dedicated PPU. I think alot of people will be surprised at what this tech could mean and we are less then a year away from having a GLquake of sorts convince the industry that this isn't a gimmick. So munky you play with you Millinea or Virge while we Physics believes will be playing with or Voodoo1's. Do I think It will be as big a deal as that was? No, but I believe it will be close.

My god man, type in complete sentences. That took 5 mintes to read and digest!

 

[Rage187]

pfft, aint worth $50

 

[xtknight]

I don't understand the logic behind comparing getting a second core for the same price as a PhysX card. The physics processing performance of one A64 2.2 GHz core isn't even close to that of a PhysX card. That's like saying that second core could play games without needing a graphics card. It's ridiculous... Based on my Quake3 benches, the Radeon 9500 PRO is 100x as fast as my old P4 2.6 GHz at rendering graphics. The chasm widens with shader-intensive games. The same goes for physics. More complex objects and the CPU just doesn't suffice, even if it operates at 100 GHz. It does not have one-cycle instructions for transforming vertices like the PhysX processor does. It most likely takes at least 100 instructions to do what the latter would do in five or less. The CPU is simply versatile, but that does not mean it's fast at everything. That just means it can do everything, but some things at a sloth's pace.

 

[tw1164]

Originally posted by: Acanthus

Originally posted by: Genx87

Originally posted by: Sudheer Anne
people don't want to buy more add-in cards, and for those with SLI setups it's likely that all their slots are unusable anyways. if this card was under $100 people might jump on it, but $299+ is a pricetag almost nobody will pay. if they keep prices where they are they are simply shooting themselves in the foot.

It really depends on the titles they have lined up. If they have compelling titles people will buy it, including myself.

My only concern however is not having enough PCIE slots but I am thinking we should start seeing more that two x16 slots on these MBs. If not this is a great opportunity for Nvidia to get their duo cards out for people who want to SLI, but also want a PPU.

3 big MMOs will support it. The hardcore players in those will almost certainly pick up the card. Im taking a serious look at warhammer because i loved daoc for its first few years before mythic destroyed it.

Shadowbane II (will probably bomb like the 1st), Warhammer Online, and Vanguard: Saga of Heroes.

City of villians supports it too. COV has been using Ageia Physx software solution (for "rag doll effects"), and I think it looks great. I'm looking foward to what they can do with the PPU.

 

[Munky]

So, if one game used the physx chip to accurately model the leaves in a tree moving in the wind, and another game uses a simple vertex shader to approximate that movement, you would be able to tell the difference? Would the difference justify the cost of the chip? I still dont see why this obsession with exact physics modeling in a game when everything else in the game is based on approximations and illusions.

 

[Acanthus]

Originally posted by: munky
So, if one game used the physx chip to accurately model the leaves in a tree moving in the wind, and another game uses a simple vertex shader to approximate that movement, you would be able to tell the difference? Would the difference justify the cost of the chip? I still dont see why this obsession with exact physics modeling in a game when everything else in the game is based on approximations and illusions.

Because the industry constantly demands better?

The only way we are going to get cinematic computing is through this path, dedicated physics processors arent a matter of why, just when.

 

[Acanthus]

Originally posted by: flashbacck
I'm leaning towards the "PhysX card will Flop" category. I'd rather have a multi-core CPU that'll be used in more programs than a dedicated card that'll be used in a few games. But to be fair we need some benches to make a good judgement. Something like:

$250 PhysX + $125 A64 3000+
vs
$355 A64 X2 4200+

(prices from RTPE)

Any games tested would have to be optimized for both the physx card and multicore cpu's, of course.

Using the same level of complexity, the 3000+ and PPU would demolish an X2. You understand the architecture right?

 

[Ctrackstar126]

People just need to get better jobs and we need to lower the hiring age to like 10 therefore kids in middle school can afford awesome systems and graphics and physics can move at a faster rate cause more people can afford it.

 

[Creig]

Originally posted by: munky
So, if one game used the physx chip to accurately model the leaves in a tree moving in the wind, and another game uses a simple vertex shader to approximate that movement, you would be able to tell the difference? Would the difference justify the cost of the chip? I still dont see why this obsession with exact physics modeling in a game when everything else in the game is based on approximations and illusions.

By allowing the PPU to calculate the effects of weapons fire on a building, you will have a different result each time you play the level instead of the exact same visuals every time. Maybe one time half the building will collapse or maybe the whole thing will go down. One time it collapsed in on itself, while next time it collapsed to the north, burying an enemy tank at the same time.

It's all about realism.

 

[Effect]

Nicely explained, realism. I like the concept behind the PPU, and i can see it becoming a big part in gaming hardware, but i think we're still a while away from that. As soon as the technology matures, and the results justify the costs, i'd love to pick up a PPU, after all, isn't realism the path gaming has been taking for....well....ever?

 

[DeathReborn]

Originally posted by: Creig

Originally posted by: munky
So, if one game used the physx chip to accurately model the leaves in a tree moving in the wind, and another game uses a simple vertex shader to approximate that movement, you would be able to tell the difference? Would the difference justify the cost of the chip? I still dont see why this obsession with exact physics modeling in a game when everything else in the game is based on approximations and illusions.

By allowing the PPU to calculate the effects of weapons fire on a building, you will have a different result each time you play the level instead of the exact same visuals every time. Maybe one time half the building will collapse or maybe the whole thing will go down. One time it collapsed in on itself, while next time it collapsed to the north, burying an enemy tank at the same time.

It's all about realism.

QFT. As far as it being a flop, it's new technology and it already has 1 game available that supports it (City of Vilains). In 5 years you might not be able to buy a new PC without a dedicated PPU or it might just fade away. Every way you look at the PPU there is a potential market for it. Scientists will be buying PPU's to help in research

Expanding on what Creig said about the buildings, thePPu would allow for accurate bullet movement upon entering water, let structures in the game behave as they really would. I'd like to see thingslike bullets from a high powered rifle going through a wall say and a low power rifle would just embed in the wall. I know the basics of that are already in place but the PPU will finally allow developers to make a game that is close to real (Rainbow Six would be great with better physics).

 

[Fox5]

Originally posted by: Genx87

Originally posted by: munky

Originally posted by: concernedsophist
I am so surprised at how many ?anti-physics people? are around here! Come on people. Obviously it will enhance game play! Even if the first generation of cards is overpriced and underutilized it is a stepping stone to bigger and better things (read cheaper and more powerful). So don?t buy the $300 card this year. However, it is a necessary step to make these cards to make advances in physics hardware, software, and market penetration. Imagine the future games where the cpu is almost entirely free to do AI! This will lead to huge advances in game play. How can you be a computer and game enthusiast and not be excited about the prospect of this technology, even if you are not going to purchase an early version of the hardware?

I'm an enthusiast, but I'm not some rich fat bastard with more money than brain cells. Ageia is in this business only for one reason - to make money. They tell you that a physx processor will make games so much more realistic because it's in their best interest. No games even fully use the two cores on my cpu, and you want me to believe that the cpu is too weak and slow to handle physics? What kind of physics does a game require to warrant a dedicated processor? Until I see a game that justifies the cost of a physx chip, I will not support it.

CPU's are general pupose processors. When was the last time you saw a general purpose processor outperform a dedicated processor?

Of course they are in the business to make money but what does that have to do with the technology behind it? Clearly games with more physics will be > than games that are void of it.

The cpu can outperform sound cards, even the XFi.

 

[Acanthus]

Originally posted by: Fox5

Originally posted by: Genx87

Originally posted by: munky

Originally posted by: concernedsophist
I am so surprised at how many ?anti-physics people? are around here! Come on people. Obviously it will enhance game play! Even if the first generation of cards is overpriced and underutilized it is a stepping stone to bigger and better things (read cheaper and more powerful). So don?t buy the $300 card this year. However, it is a necessary step to make these cards to make advances in physics hardware, software, and market penetration. Imagine the future games where the cpu is almost entirely free to do AI! This will lead to huge advances in game play. How can you be a computer and game enthusiast and not be excited about the prospect of this technology, even if you are not going to purchase an early version of the hardware?

I'm an enthusiast, but I'm not some rich fat bastard with more money than brain cells. Ageia is in this business only for one reason - to make money. They tell you that a physx processor will make games so much more realistic because it's in their best interest. No games even fully use the two cores on my cpu, and you want me to believe that the cpu is too weak and slow to handle physics? What kind of physics does a game require to warrant a dedicated processor? Until I see a game that justifies the cost of a physx chip, I will not support it.

CPU's are general pupose processors. When was the last time you saw a general purpose processor outperform a dedicated processor?

Of course they are in the business to make money but what does that have to do with the technology behind it? Clearly games with more physics will be > than games that are void of it.

The cpu can outperform sound cards, even the XFi.

Right, but having dedicated hardware dramatically reduces overhead.

 

[Fox5]

Originally posted by: Acanthus

Originally posted by: Fox5

Originally posted by: Genx87

Originally posted by: munky

Originally posted by: concernedsophist
I am so surprised at how many ?anti-physics people? are around here! Come on people. Obviously it will enhance game play! Even if the first generation of cards is overpriced and underutilized it is a stepping stone to bigger and better things (read cheaper and more powerful). So don?t buy the $300 card this year. However, it is a necessary step to make these cards to make advances in physics hardware, software, and market penetration. Imagine the future games where the cpu is almost entirely free to do AI! This will lead to huge advances in game play. How can you be a computer and game enthusiast and not be excited about the prospect of this technology, even if you are not going to purchase an early version of the hardware?

I'm an enthusiast, but I'm not some rich fat bastard with more money than brain cells. Ageia is in this business only for one reason - to make money. They tell you that a physx processor will make games so much more realistic because it's in their best interest. No games even fully use the two cores on my cpu, and you want me to believe that the cpu is too weak and slow to handle physics? What kind of physics does a game require to warrant a dedicated processor? Until I see a game that justifies the cost of a physx chip, I will not support it.

CPU's are general pupose processors. When was the last time you saw a general purpose processor outperform a dedicated processor?

Of course they are in the business to make money but what does that have to do with the technology behind it? Clearly games with more physics will be > than games that are void of it.

The cpu can outperform sound cards, even the XFi.

Right, but having dedicated hardware dramatically reduces overhead.

Of course, though too bad it needs dedicated ram because the pci bus is too slow. But yeah, unless you have dual core trying to do sound on the cpu could have a major performance hit....except most cpus now are capable of well over 100fps in many games, which leaves plenty of power to run those game's sound engines. (though if a game existed that made full use of the xfi's capabilities it would be infeasible to run it on a single core)