Fredrik Kjolstad
|
PhD Candidate Computer Science and Artificial Intelligence Laboratory Department of Electrical Engineering and Computer Science Massachusetts Institute of Technology fred@csail.mit.edu Curriculum Vitae |
I aim to make it easy to program sparse computing applications that can run everywhere with end-to-end performance that matches or exceeds the best hand-optimized codes. A sparse application computes on a system where most parts do not interact—they are sparsely interconnected. Examples include sparse matrices, meshes, and graphical arrangements used to simulate the dynamics of robots, optimize the flow of traffic, compute the rank of web pages, estimate the similarity of genome loci, or train sparse deep neural networks. Sparse computing is ubiquitous and will become increasingly important with the growth of robotic and autonomous vehicle applications, data analytics, and machine learning.
Throughout my PhD studies, I have developed the taco compiler for computing sparse tensor expressions and the Simit programming language for computing on sparse systems. My broad research interests include compilers, programming models and languages, parallel computing, software engineering, and performance engineering.
News
I am co-organizing an Invited Workshop on Compiler Techniques for Sparse Tensor Algebra that will bring together researchers from twenty universities, companies, and national labs to discuss approaches to sparse tensor computation, how they relate, and how ideas can be combined.
Publications
|
Sparse Tensor Algebra Optimization with Workspaces Fredrik Kjolstad, Peter Ahrens, Shoaib Kamil, and Saman Amarasinghe International Symposium on Code Generation and Optimization (to appear), February 2019 |
|
|
Format Abstraction for Sparse Tensor Algebra Compilers Stephen Chou, Fredrik Kjolstad, and Saman Amarasinghe Proceedings of the ACM on Programming Languages, Volume 2, Issue OOPSLA, October 2018 |
|
|
Taco: compilation et génération de code d’expressions tensorielles David Lugato, Fredrik Kjolstad, Stephen Chou, Saman Amarasinghe, and Shoaib Kamil AVANCÉES, Volume 12, 2018 |
|
|
taco: A Tool to Generate Tensor Algebra Kernels Fredrik Kjolstad, Stephen Chou, David Lugato, Shoaib Kamil, and Saman Amarasinghe 32th IEEE/ACM International Conference on Automated Software Engineering (tools paper), November 2017 |
|
|
The Tensor Algebra Compiler Fredrik Kjolstad, Shoaib Kamil, Stephen Chou, David Lugato, and Saman Amarasinghe Proceedings of the ACM on Programming Languages, Volume 1, Issue OOPSLA, October 2017 Distinguished Paper Award |
|
|
Simit: A Language for Physical Simulation Fredrik Kjolstad, Shoaib Kamil, Jonathan Ragan-Kelley, David I. W. Levin, Shinjiro Sueda, Desai Chen, Etienne Vouga, Danny M. Kaufman, Gurtej Kanwar, Wojciech Matusik, and Saman Amarasinghe ACM Transactions on Graphics (presented at SIGGRAPH 2016), May 2016 |
|
|
Why New Programming Languages for Simulation? Gilbert Louis Bernstein and Fredrik Kjolstad ACM Transactions on Graphics (perspective), May 2016 |
|
|
MPI Datatype Processing using Runtime Compilation Timo Schneider, Fredrik Kjolstad, and Torsten Hoefler 20th European MPI Users' Group Meeting, September 2013 Best Paper Award |
|
|
Automatic Datatype Generation and Optimization Fredrik Kjolstad, Torsten Hoefler, and Marc Snir 17th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming (short paper), February 2012 |
|
|
Transformation for Class Immutability Fredrik Kjolstad, Danny Dig, Gabriel Acevedo, and Marc Snir 33rd International Conference on Software Engineering, May 2011 |
|
|
Bringing the HPC Programmer's IDE into the 21st Century through Refactoring Fredrik Kjolstad, Danny Dig, and Marc Snir SPLASH 2010 Workshop on Concurrency for the Application Programmer, October 2010 |
|
|
Ghost Cell Pattern Fredrik Kjolstad and Marc Snir 2nd Annual Workshop on Parallel Programming Patterns, March 2010 |
|