For a brief moment, it once looked like hardware-accelerated gaming physics had the potential to transform PC gaming's static rooms full of storage crates into dynamic worlds full of showering particles, rumpling cloth and realistically collapsing bridges. Co-founded by Manju Hegde, a new start-up called Ageia ambitiously launched its dedicated PhysX accelerator; a technology that's now a part of Nvidia's family, and it looked like gaming physics was about to completely overhauled.
New report from Skillsearch found that 22% of those surveyed had been laid off within the past 12 months.
It's a step forward for Stop Killing Games.
The Callisto Protocol director thinks the solution involves the right people, the right timing, and perhaps a little bit of AI
I don't agree with that. I WISH I could agree with that. But buying habits and customer opinions prove otherwise
We've seen developers in the AAA space try new things and ideas. More often than not, the customers aren't willing to give things a chance, or not enough people buy into the project for it to grow.
Creativity works better in the indie space because the budgets, pressures, and expectations aren't the same.
it's a nice idea and it worked during the PS2/PS3-era when AAA didn't cost hundreds of millions of dollars. smaller budgets and shorter development time left room for more creativity and more risk. a game didn't need to sell 4 million+ copies to break even. things are different now.
Honestly, complex physics is pretty invisible to the typical gamer. Accelerating it in hardware isn't all that useful, because the real expense of physics comes with sorting the bazillion objects you want to have in the world (sorting is NEVER cheap), not doing the accelerated precise collision mechanics.
About the only thing hardware physics gives us, is the ability to run a fancier simulation on a zillion small, simple objects, in a restrained space against a few more complex geometrical objects. That way you don't need to sift through piles of data to find what you want to collide against, and hardware can do its thing -- be fast without memory issues getting in the way.
In short, about the only things hardware physics are useful for are things like more accurate smoke, sparks, etc simulations. Constrained, compound objects like ragdolls become exponentially expensive as you attempt to make them more complicated, solve motor chains, etc. While it seems clever to throw hardware at this, really the end result is that you end up with wildly variable hardware physics performance, and you just end up having to limit the physics you use anyway, so you don't have performance spikes.
It's not like graphics. It just isn't. Physics is too variable / unpredictable to rely upon hardware acceleration to yield much improvement, relative to its cost. Game devs keep it simple because it not only performs better (HW acceleration would improve performance), but also because it just plain works better when its simple (hardware acceleration would not help make simulations more robust, relative to their complexity).
You can model a bicycle as a box, or low-poly mesh, and wrap an 11-body ragdoll around it, and have it be fast, and work. Or you could model it as a 10000-polygon mesh, and wrap a 45-body ragdoll around it, and have it be fast (with HW accel), and totally whacked out and busted (no matter what you do).