[ODE] Quick, quickstep comparision

[Billy Zelsnack]

Megan Fox wrote:

>Impressive - so at least now ODE is a far sight better than Newton.
>
>
>
>
>
>That said, can anyone say why Tokamak performs consistently better? 
Is it
>
>their algorithms being superior, or is this just a case of quickstep
still
>
>being a little unoptimized and generally new code?
>
>  
>

It is very difficult to fairly compare. Tokamak seems to handle large 
mass ratios better and Newton handles large mass ratios much better. 
Quickstep seems to be able to handle around 20:1 where Newton can
handle 
200:1. Still who knows what tricks the other engines are using. (The 
Newton engine may very well not be using any).

BTW. For large mass ratios, last years siggraph paper "Rigid bodies
with 
stacking" (or something like that) has a simple technique that works 
pretty well.

Considering how new this code is, it works amazingly well already.
There 
are tons of tricks that can be done and this new code is much more 
approachable than the previous solvers. I'd expect quite a few more 
people tinkering with it or creating variants.




	
		
__________________________________
Do you Yahoo!?
SBC Yahoo! - Internet access at a great low price.
http://promo.yahoo.com/sbc/