About Me

Reading time ~1 minute

Hi! My name is Cade Brown. I like to do programming, maths, and music.

My interests are: numerical computation, programming language design/theory (PLT), and audio/video processing.


Here are my best/popular/favorite projects to date:

“Know It All”, an A.I. writer

Demonstration of “Know It All”, an A.I. writer

  • This was my, along with Jakob Liggett’s, submission to VolHacks IV, in which an AI model trained on internet text posts “completes” your thoughts.

You can see the (currently in a somewhat-messy state) source here.

SimpleSummit, a distributed computing project

  • Check out our website, here: https://simplesummit.github.io/
  • I did all the software for this, which included writing a fractal explorer which ran on 8 different machines at once, which communicated in real time over a network to produce an image around 30 frames per second.
  • This was a main project I worked on during my internship at Oak Ridge National Laboratory (ORNL), for my 11th and 12th grade summers.

Check out or ORNL article here: SimpleSummit

MAGMA (Matrix Algebra on GPU and Many-core Architectures)

  • This is my current (as of Jan 2020) project, which I have been given the task of porting to AMD GPUs (right now, it relies on NVIDIA).
  • This is my largest project, in terms of reach, and will be a large part of the Frontier HPC system, which will only have AMD GPUs.

See MAGMA for more information


Additionally, for reference, here is my (approximate) employment record:

Intern @ ORNL (Summer 2016 - Summer 2017)

For 2 summers, I worked as an intern at ORNL (Oak Ridge National Laboratory). There, I helped write SimpleSummit (https://simplesummit.github.io/), including the fractal rendering program, the website

Research Assistant @ ICL (Summer 2019 - Present)

Since I’ve started attending UTK, I’ve worked at ICL (Innovative Computing Laboratory). Within ICL, I work under Dr. Stan Tomov in the Linear Algebra (LA) division. Our task has been porting and accelerating common LA operations on the new AMD HIP/ROCm stack.