Becoming a physics programmer

Started by
39 comments, last by GameDev.net 18 years, 5 months ago
Quote:Original post by kobe bryant
Thanx for the feedback, how long will it take to learn all that math and what are the best books to learn these things.


If I may, I'd like to second this question as this is an area I'm also very interested in.

Cheers for all the information posted.
Advertisement
I don't know if this is a field of work that I would really want to get into I'm sure that as soon as physics chips come out a lot of the physics programming jobs are going to be relegated to API calls. With the physics being hardcoded into the chips.
Quote:Original post by Vanke
I don't know if this is a field of work that I would really want to get into I'm sure that as soon as physics chips come out a lot of the physics programming jobs are going to be relegated to API calls. With the physics being hardcoded into the chips.


Hmmm.. I can imagine something very similar being said a few years back...

"I don't know if this graphics programming is a field of work that I would really want to get into. I'm sure that as soon as 3D graphics chips come out a lot of the graphics programming jobs are going to be relagated to API calls. With the graphics being hardcoded into the chips."

The point being that whilst it will get less and less necessary as time goes on to be able to implement low level physics from scratch, there will always be a need for people who know how to take either their own or someone elses (software or hardware) physics engine, and make it work within the considerable constraints that a game imposes. To do this well will always require a deep understanding of the principles and theories involved, as well a practicle experience.
Quote:Original post by kobe bryant
Thanx for the feedback, how long will it take to learn all that math and what are the best books to learn these things?



I'm not entirely certain what the best books to learn the math are, but I've been learning linear algebra and tensor math from books I got at a library downtown. I'd say about 30% of the books I've encountered are good enough to learn from and since there are atleast 200 I can get for free... My point is that if you have a library near by you should definitly check there before you spend a lot of money. As far as time, that probably depends on the person, but odds are you can learn much faster than you would at school. Since about April I've read a book on basic calculus, graph theory, linear algebra, less basic calculus, and right now I've got a book out on tensor math and another one on general topology.


What exactly is topology used for in programming? Every other book I got out becuase I wanted to program something, but I've never needed topology...
Quote:Original post by Anonymous Poster
Too many people download code off the internet and never understand what it's doing



Thats one thing I hate doing. I like to get code for learning purposes, but I DON'T like to use it until I understand exactly how it works. And since I'm not a very good self teacher.. I get stuck often :/
If you gave a helpful reply, I rated you up.
Topology teaches you the basis relation beetween metric spaces, if you consider an object like a volumetric metric space, you could end up with interesting algorithm ( mostly based on discret division ) for doing collision detection or interpenetration, also consider that you could apply the same algorithm for visibility detection purposes.
This is a huge topic that i haven't seen covered anywhere in game programming, with time , and computing power, many of those algorithm will become more and more popular.
Hi i've just start a-levels in maths, physics and electronics, what would you sugest is a good cource to start with in computer programming, i would like to specialise in physics coz i'm very interested in it, but would that mean do maths, physics and a computer programming degree?

ps I'm currently self-teaching myself C++, have nearly finished pong with allegro.
I have a feeling that the Ageia physics chip might not be a fixed function pipeline. It might have a programmable pipeline that is analogous to the pixel and vertex shader pipelines in modern GPU's. Therefore, working with physics hardware could become an extremely interesting thing to do!
Graham Rhodes Moderator, Math & Physics forum @ gamedev.net
Just as with graphics, a fixed, high-level physics API will become a hindrance for companies that want to innovate. Thus we have programmable shaders on GPUs. It remains to be seen if the PPU manufacturer(s) will open things up.

I have had to add many low level features to my physics engine to support our game. This would have been difficult or impossible to accomplish with a third party engine.
Erinhttp://gphysics.com
Quote:Original post by kobe bryant
I made this post specifically for people who are physics programmers or people who know about the ins and outs of being one in the games industry.

I decided recently that physics programming was the area of games programming i would like to persue, i have no formal training in physics, i didn't do it at degree or college (high school for the yanks) level. I am currently doing my masters degree in games programming, we will do a physics module next term. My supervisor recently informed me that programmers without degrees or PHD's in maths/physics are very unlikley to be employed as physic programmers.

I came here to get some more opinions, i dont expect a games company ask me to write a physics engine after one semester of rigid body physics but i did feel that i could acquire the knowledge over some years at a company and specialise in this area eventually.

Any advice or opinions will be very welcomed.


Wanna hear a funny story?
I'm a 13 years old russian boy specializing in computer graphics and algorithms.
I know C and VB.
I wrote my own 3D engine from a scratch.
I know OpenGL.
I got excelent grades in chemistry and physics and know many things on the subjects.
I wrote tons of really cool demos.

"you're obviously very intelligent in some subjects, including maths. I've got a first class in CS (though not from as good a uni as Thomas) and I know very little in comparison to you and Thomas"

That's a quote of a conversation I had in another forum, in a thread about collision detection.

But i'm not writing this to brag, but to show you how bad life is when you are good at something and don't have a proof for it, so read on:

I study in weakest math group (as a result of a long series of extremely stupid mistakes, not math mistakes, life mistakes) and even my own mom thinks i'm a failure and doesn't even want to take a look at the stuff i'm doing!
So how can I expect a director of a computer games company to look at them?

Now isn't that totally ironic?
I could use some advice as well...

This topic is closed to new replies.

Advertisement