Leaving this here as an interesting paper (which I have by no means finished reading) on reducing PageRank (and other sparse algorithms) to a sparse Linear Algebra operations. It is also notable for pointing out the existence on a framework which may already parallelize operations.
(I came across this while digging on the Matrix multiplication article you had).
Leaving this here as an interesting paper (which I have by no means finished reading) on reducing PageRank (and other sparse algorithms) to a sparse Linear Algebra operations. It is also notable for pointing out the existence on a framework which may already parallelize operations. (I came across this while digging on the Matrix multiplication article you had).
Interesting implementation of Pagerank for GPUs https://par.nsf.gov/servlets/purl/10294352
Mostly I am just documenting this as an interesting avenue more so than any specific action item.