1. Not finding help here? Sign up for a free 30min tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Computational Science

  1. Jan 27, 2006 #1
    What kind of jobs would someone focused in computational science (or more specifically comp. physics) be able to fill?
  2. jcsd
  3. Jan 28, 2006 #2
    depends on what you can do...main focuses should be in numerical methods & Optimization(numerical optimization) and 3D rendering/visualizatoin in science(realtime/high performance/scene management techniques)

    AI is a fun computational science. so is 3D physics based rendering(planetary or cosmology is my preference or where i want to go before jumping into AI) though i can't deny that rigid body stuff is fun but thats more for gaming.

    [0]research in optimization and fast-paced scene management techniques
    [1]You could build science based virtual reality environments...
    [2]or build state-of the-art physics based RTS/FPS games

    and as the last resort...go get your BEd. and become a awe-inspiring highschool teacher. And make the next generation of programming students.
    Last edited: Jan 28, 2006
  4. Jan 28, 2006 #3
    The problems with simple analytical solutions have been solved ages ago and therefore computational physicists are present at all subfields of physics. Landing a job probably won't be an issue. In condensed matter physics you'd be sought after to do band structure calculations and simulations of measurements and so on.
  5. Jan 28, 2006 #4
    I assume nanotechnology is no exception? :smile:

    Is this assuming I've done a PhD or at least a masters degree?

    Lastly, I see some universities offering computation science undergrad degrees (like UW). Is this the normal route to go, or do most get a BSc in physics then do a PhD in compsci?
    Last edited: Jan 28, 2006
  6. Jan 28, 2006 #5


    User Avatar
    Science Advisor
    Gold Member

    Certainly not, comp phys and can go anywhere in physics and 'suffer' from too many options & job offers :wink: .
  7. Jan 28, 2006 #6

    I am currently doing my Physics PhD on computational physics in nanotechnology. More specifically, i do DFT (density functional theory) calculations onto the interface between the metal gate and a high k material (ie a material of which the dielectric constant exceeds 3.9 times the vacuum value) of a MOSFET structure. The reason for this research is trying to quantify the interactions in the interrfacial region that govern the interface electrostatics.

    This is very necessary since both metal gates and high k materials (like HfO2 or HfSiO4) will be needed to further scale down the dimensions of a MOSFET, thereby following Moore's Law. My calculations need to provide more info on "which experiments" to do or what processes of the MOSFET engineering (like deposition, defect creation, etc) influnce the interface electrostatics in a "bad way". A "bad way" means "harming transistor operation and thereby blocking the miniaturization". Hence, i work very closely with experimentalists and engineers, though i was trained as a theoretical physicist (QFT major) at college.

    This is just one example of how computational physics can be used in applied science and i can assure you that lots of companies like Intel, Panasonic, ASM, Philips (some of our core partners at IMEC) are interested in the application of computational physics into their research departments because doing such calculations reduces production costs.

    CERN is also very much investing in computational physics for their research purposes, just to give you another example.

    In short, learn as much as you can (many-body physics : Hartree Fock and DFT; software to use like : SIESTA, CASTEP, ABINIT, etc) because you can do many things with it. Just google to some of these names to find out more. Once you have a PhD in this field (coming from a respectable institution or college) your options are wide open in both research and development departments. People can really benefit from your skills as a computational physicist.

    Last edited: Jan 28, 2006
  8. Jan 28, 2006 #7
    Marlon, much thanks for this response. So you major'd in QFT? Does this mean at the undergraduate level?

    What is the major a student would seek at the undergrad level if one was to do computation physics?
  9. Jan 29, 2006 #8
    Here in Belgium and the rest of Western Europe, one needs to study five years in college to acquire a master's degree in physics. I my time (three years ago, that was four years in stead of five). QFT is covered in the fourth and fifth year. I did my master's thesis on quarkconfinement in QCD. One must have the master degree if you wanna persue a PhD. The bachelor degree (first three years) is not worth anything because most companies will hire masters or PhD's to do the "serious" work. A bachelor mostely will end up in a situation where he/she find him/herself doing the job a dressed monkey can do too:wink:

    Major in condensed matter physics, many body physics (Hartree Fock, DFT, Monte Carlo simulations,...), those are the real fields on interest. One can approach these fields both from a more theoretical level (theory behind solving the many body Schrödinger equation) or from a more practical level (where you chose to work with the software directly and learn about the theory as you move on). Both ways are good since most of the basic theory will be covered in your solid state course.

  10. Jan 29, 2006 #9
    Thanks marlon!
  11. Jan 29, 2006 #10

    Dr Transport

    User Avatar
    Science Advisor
    Gold Member

    Industry is using computational scientists more and more during the design phase. I work in the aerospace industry, each and every test we perform on a sub-system must be proceeded by a prediction of outcome. We spend alot of time after the fact doing computational analysis on the data re-modeling the systems to get a better handle on ALL the physics involved and generalizing our codes. Sometimes we get it right initially, sometimes after a bunch of work and sometimes never. Sometimes the measurements are wrong and you never find the problem computationally.
  12. Jan 29, 2006 #11
    cscott if you want to do reserach your best bet is to go for an MSc in physics and a BSc in CS(computational science OR computer science & MATH)/minor in Phys.
    However if your just looking for some job like in game physics...what you need to do is create demos. Generate as much code as possible to show that you know hardware and software. This can be achieved during your undergrad year. but most universities won't guide you through it...I suggest taking a look at a visualization package(ie VTK/PTC/ OpenSceneGraph) or a 3D open source game engine(Ogre3D, Irrlicht) and numerical packages like CLAPACK, LAPACK. learn how their organized and then begin to use what you learn in and out of school to emulate them.
  13. Jan 29, 2006 #12

    Dr Transport

    User Avatar
    Science Advisor
    Gold Member

  14. Jan 29, 2006 #13
    How does that compare to doing my BSc in physics with a minor in CS and taking many scientific programming related courses? Is this feasible?
    Last edited: Jan 29, 2006
  15. Mar 31, 2010 #14

    hey..i have done M.Sc in physics and a B.Sc in computers..n frmlast 2 years i am working as a software developer in an IT company..hw can i come in this computational science field??
  16. Mar 31, 2010 #15


    User Avatar
    Gold Member

    would it be more beneficial to do the undergrad comp phys degree then move on to comp phys Ph.D, or obtain a degree in physics during undergrad then declare for comp phys for Ph.D.?
  17. Apr 1, 2010 #16


    User Avatar
    Gold Member

Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?

Similar Discussions: Computational Science
  1. Computational Science (Replies: 6)

  2. Computer Science (Replies: 3)

  3. Computational Science (Replies: 3)

  4. Computer Science (Replies: 1)