Ceze, Luis HNelson, Jacob Eric2015-02-242015-02-242015-02-242014Nelson_washington_0250E_13858.pdfhttp://hdl.handle.net/1773/27449Thesis (Ph.D.)--University of Washington, 2014Grappa is a modern take on software distributed shared memory (DSM) for in-memory data-intensive applications. Grappa enables users to program a cluster as if it were a single, large, non-uniform memory access (NUMA) machine. Performance scales up even for applications that have poor locality and input-dependent load distribution. Grappa addresses deficiencies of previous DSM systems by exploiting application parallelism, trading off latency for throughput. We evaluate Grappa with an in-memory map/reduce framework (10x faster than Spark); a vertex-centric framework inspired by GraphLab (1.33x faster than native GraphLab); and a relational query execution engine (12.5x faster than Shark). All these frameworks required only 60-690 lines of Grappa code.application/pdfen-USCopyright is held by the individual authors.graph; irregular; latency; network; programming; throughputComputer scienceComputer engineeringcomputer science and engineeringThesis