Project 2: Matt Elser

[mattelser]

Repo Link

• implemented all required algorithms (cpu, gpu, and thrust implementations of scan and compaction)
• Efficient GPU implementation is faster than CPU for arrays of size 2^17 and greater
• compaction/scan works up to arrays of size 2^28, after which the test machine runs OOM
• implemented radix sort, but without tiling and bitonic merge

[mattelser]

I spoke with Prof. Mohammed after class and was told I could update the readme so I did