Jump to content

sjr

Member
  • Posts

    8
  • Joined

  • Last visited

Awards

This user doesn't have any awards

About sjr

  • Birthday February 22

Profile Information

  • Gender
    Male
  • Location
    Cambridge, Massachusetts
  • Interests
    High Performance Computing, Machine Learning, Physics Simulators, General Tech Enthusiast
  • Biography
    B.Eng and M.Eng (Mechanical Engineering and Computer Science)

    Currently doing my PhD at MIT in Computational Science and Engineering

    Write simulation software to model solid, liquids...anything really...
  • Occupation
    PhD Student - MIT

Recent Profile Visitors

331 profile views
  1. So from personal experience with Maya you'll want at least 64GB DDR4 and an SSD. Some of the features in those programs don't utilise the GPU so instead of just focussing the money solely on GPU balance the spend on CPU and GPU.
  2. Hey fellow programmers, Fairly new to the Forums so just wanted to say 'hey'. I'm currently doing my PhD at MIT in Computational Science and Engineering. I write computational mechanics software mostly, so basically PDE solvers using something called SPH (and MPM). We work on making some of the physics look half decent (image* below if it works) but we still have a ways to go. Backend of the code is written in C++ but the spirit of C, very few objects. Fully parallelized on CPU and working on CUDA acceleration as well (though having some issues with the way we want to try it). I guess I'll just list what I use: - VSCode/Nano/VisualStudio - CMAKE - C++/Python - OpenMP - CUDA - SWIG (really cool way to make a python library from the C++ code and lets users write python scripts, fast and reasonably user friendly) - Paraview & Houdini (for visualisations) All of it is compiled for Linux (where our super computers run [specifically Ubuntu and CentOS]), macOS (coz...umm okay), and Windows That's mostly what I work on for my research but have been playing around with Node.js, AWS, JavaScript, and just various stacks. But yeah happy to chat about any/all of this stuff. Looking forward to being a part of these forums! Sam.
  3. He probably meant this one, was pretty.................... cool It's a bit more like...the state of the bit is undetermined until you put it through certain Quantum Logic Gates. You can build most of the same XOR's, XANDs, etc. but the neat thing with QuBits (Quantum Bits) is that you can tie the results of particular logical evaluations together in a long string of entangled QuBits and then read it once and get the probabilistic result. Oh also, Qubits take up a lot of memory so most Quantum Computers can hold something like 16 QuBits max....which is like 1 double-precision integer in terms of normal bits. It's partly how the Shor's algorithm for breaking RSA works, they can get the factors of the large number and then put that into a standard computer and can spit out the actual keys. There's some great walkthrough videos that deal with it. I'm trying to see if there are some applications for my research in speeding up some of the O(Nlog(N)) algorithms to like O(N) or better! But yeah nerding out has a time and place.
×