Faster Matrix Algebra for ATLAS

Eigen is a C++ template library for linear algebra that aims for high performance in combination of high reliablity and good compiler support. A lot of remarkable projects rely on it, including Google's Tensorflow. Another successful project using Eigen is the high-energy physics experiment ATLAS at the LHC.

At the LHC, millions of particles collide every second and each collision creates a huge amount of data that has to be classified and analyzed by software. Most algorithms in ATLAS software use symmetric matrices, i.e., matrices where the upper triangular part is equal to the lower triangular part. Unfortunalty Eigen currently misses support for symmetric matrices.

This Google Summer of Code 2018 project aims to implement a class for handling symmetric matrices in Eigen. The goal is to provide a working implementation that can be submitted as a patch for Eigen.

This project proposal contains implementaions ideas and plans, a detailed implementation timeline, consisting of 13 weekly tasks and some short biographical information about me.