optional physics more better "ageia"

Quote:

---

How do I become a licensed developer and what are the licensing terms?
Costs
At present, the following two options are all that are available. We will likely never provide any sort of royalty-based license. Open-source usage is limited only to the most middleware-friendly license model--in other words, we retain full ownership and rights to our own IP, and what you ship of ours will still be specified by contract.

Free:
Non-commercial use
PS3 platform (through Sony pre-purchase)
Through some of our middleware partnerships, such as UE3, Gamebryo 2.2, and others--often limited to non-commercial use
PC platform (if the game makes significant use of PhysX HW)
$50k per platform:
All other uses
Fee may be waived at our disgression for multi-platform developers providing PC HW support
Fee may be waived at our disgression for some Tools & Middleware providers

---

Quote:

---

However, disabling the requirement of having physics through a commandline would be the next best option.

---

Quote:

---

I've heard that Microsoft is delevopping "Direct Physics"

---

Where did you hear this? If it is true, I'd like to know more about it.
Some people (me included ) have talked about DirectPhysics but, AFAIK, Microsoft hasn't done anything.

Quote:

---

I still dont think the future of games/simulations is a specialized hardware for everything.

---

Quote:

---

I dont see why CPU-based physics isnt good enough for the future... As CPUs get faster, won't physics models improve as well?

---

True. But why do we still spend $80 to $800 on special video cards? And why is sound still handled by a separate chip (which is now part of the motherboard)? CPUs today could easily handle the graphics you had in Quake 2 but what hardcore gamer would be happy today with that when they could have shaders, mirrors, blur, bump-map, etc. all at 90+ FPS?

If (and this is a big IF) physics can benefit the same way as graphics and sound does by having special hardware, then what hardcore gamer would be happy with HL2 physics when they can have a 1000+ objects on the screen all moving with realist forces; water that can "pour" out of pipes, onto desks, around the player's legs, causing all the objects in the room to float depending on their mass; and car crashes that are not "canned" but actually model the damage to the vehicle (from dents and cracked tail-light up to bending the frame and cracking the engine block). These are the sort of hype that people are saying physics hardware can deliver.

Now we have to see if reality lives up to the hype.

Quote:

---

but this chip isn't that special. shader pipelines (especially shader model > 3) can do the same. there is no reason to add this physics chip anywhere. the reviews and sales so far also have been very poor.

ati and nvidia will just have to add more and more shader pipelines which will happen anyway (the top cards already have around 48). then a bunch of them always can be used for physics.

---

Quote:

---

The same could happen with physics chips. They get incorporated into graphic cards or mother boards. Some "freaks" will buy extra cards but not the standard user.

---

This is a more likely solution, although it will be harder to justify a physics chip on the mother board of a "normal" PC. Windows uses sound for Word, Vista uses 3D graphics, but unless Vista-2 requires physics to move the icons around, I think physics chips will probably be part of high-end graphics cards.

Quote:

---

But how easy would it be to integrate such hardware gpu phyisics into ODE, Newton or other affordable solutions?

---

Good thinking. Newton is closed, but ODE is an open source solution. You can check on their website, I would be surprised if somebody isn't working on a ODE on GPU solution (if not, you should start one ).

I'd really like to see ODE take off. There needs to be more open source tools and libraries for game development if indies want to stay "in the game". I also don't like putting too much effort into a solution that may go away overnight. If PhysX doesn't have some major success in the next year, they will be quickly forgotten.