Any thoughts on the AGEIA PhysX 100M?

Just wondering what others think about this card. From what I've read, it's very restrictive in it's use. You have to have a game that supports it directly, and if doing multiplayer, other players must also have the same card.
Definitely not knocking anyone who has or will get an M1730, but just wondering what you think.
I personally wonder, if it's just a novelty item, a gimmick. I don't think it will last long. Do you think it will catch on?

 

Until AGEIA gets more companies to tag along, those will be my thoughts. Currently, the cards only offer more eyecandy. In reviews of GRAW, it's been stated that the card can actually still lower framerates.

 

With the advent of quad and eight core processors, you'd think it would make more sense to just utilize one of those cores solely for physics processing. Probably wouldn't be as effective as a dedicated physics card, and would no doubt be a ***** to program, but I can see it happening in the future.

 

Physics on multicore is definitely the way to go since in the end it'll reach a larger audience than an add in card. It's also the way to go now that Intel has bought Havok.

I believe UT3 is going to support PhysX too, but even then I think they're classifying it as a mod that may well be incompatible with no PhysX users preventing networked play.

 

It's a waste of time and money. One day, it is possible that someone might create physics accelerators that actually do their job reliably, but it hasn't happened yet.

 

The unreal engine 3 physics engine is supposed to be based on AGEIA physics, the unreal engine is used in a fair few games and it should lead to some good games supporting the physx card

 

A physics card is great but for $300 forget it. $100 maybe. But as others have stated, with multi-core CPU's commonplace, utilize one of the cores. All they really need to do is take a good physics routine and run those calculations dedicated to one of the cores. However, due to the heavy floating point useage, GPU's may be better dedicated to this task.

It would make sense to have a multi-core GPU and have one core dedicated to physics, then it could feed the data realtime to the rendering engine.

 

Why?????????

 

Enhanced physics in a game would alter the actual gameplay, unlike graphical options that simply make things look nicer.

If say, you have a building explode + PhysX and several of the sheet metal walls come crashing onto your Jeep, causing it to explode, the dude with the building exploding - PhysX who sees nothing more than a big fireball followed by a random Jeep explosion is going to be very confused.

Having players without the PhysX card play with players that do have it would cause several incongruities in gameplay environment.

 

Couldn't they implement fancy looking physics that don't actually affect gameplay? Like if the environment was destructible it would just look better when it blows up on the PhysX card owner's computer.

 

If you are concerned about the physics in multi-player games, I must say that the PhysX would most likely not be a must since it's not a warranty for the game being in sync in all events on all computers - even completely identical computers for clients can't warranty it (like the explosions looking alike in all machines - it can affect the gameplay - like moving crates and such - and we can't have a client being different from the others, but the output of the identical machines can be different, say because of a small lag or a whole load of other things).
I think for multi-player, the server would have to tell what exactly happens in the screen and the clients would only render it, so PhysX might have no impact (except on the server) in multiplayer games.

 

It's not a gimmik me thinks. I mean AGEIA have been around on desktop for a while now. Maybe they saw the 100M as a chance to get more systems with the technology?

For the moment the games that support it seem limited, but the results with games that do use it seem to be quite nice. I mean people say "its just more eyecandy" but thats why we buy better graphics cards, for more eye candy. In this case its eyecandy that actually has an effect on gameplay, even if small.

It's good that the 100M turned up, because the more the product is out there the more support its going to get. You need to have enough of the product in place to get proper game-wide support. Hopefully since the UT3 engine makes use of it, games based on that engine will also encorporate elements in. Once games get rolling and start offering more, people will start buying into the product. Oh and once the prices come down, still too high (talking about desktop wise).

Not sure how much extra $$/££ the 100M adds to the base cost of the XPS M1730, but its nice to have.

 

I think it's standard on the 1730, IIRC.

 

For online play it would be like playing catch with a person that had no depth perception.

 

Yeah it is, and you cant "unselect" it (at least here) so i was wondering how much extra cost they are forcing on you I heard in some places that the card was actually an extra

 

It costs an extra RMB1799.46 which is roughly equivalent to US$240.23 at today's forex rate rounded up to the nearest cent. Extra costs will vary depending on region so the curious can check local dell websites and find out for themselves.