Charla: DASL: A Scala-based DSL for Graph Analytics on GPUs


Compartir
Charlista: 
Olaf Haring Postdoctoral research fellow, Hasso Plattner Institute
Fecha: 
23 Marzo, 2016 - 12:00
Sala: 
Auditorio Philippe “Algorithmix” Flajolet, Piso 3, BP
Organización: 
Prof. Jorge Pérez

Abstract:

For data intensive analytic challenges, memory bandwidth, not processor speed, is the primary performance limitation. Graphics processing units (GPUs) provide superior bandwidth to main memory and can deliver significant speedups over CPUs. However, it is not trivial to develop GPU accelerated graph algorithms. In contrast, to scale applications onto multicore, parallel architectures it requires significant expertise, including intimate knowledge of the CPU and GPU memory systems, and detailed knowledge of a GPU programming framework such as OpenCL or CUDA. To enable analytic experts to implement complex graph applications that efficiently run on GPUs we have developed a domain-specific language called DASL and a corresponding execution system that executes DASL programs by automatic translation into program code that is optimized for GPUs. In the talk I will introduce DASL based on a few simple examples, provide a technical overview on Blazegraph DASL, and discuss some aspects of the approach in more depth, including the translation of DASL statements and the integration of user defined functions.

 

Bio:

Olaf currently works as a postdoctoral research fellow in the Semantic Technologies Group at the Hasso Plattner Institute. He received his Ph.D. in Computer Science from the Humboldt University in Berlin and his Ph.D. dissertation, titled "Querying a Web of Linked Data: Foundations and Query Execution," was honored with the 2015 SWSA Distinguished Dissertation Award of the Semantic Web Science Association. Olaf's research interests are in different aspects of Web data management, such as Web query processing, query languages for Web queries, Web data provenance and quality, RDF databases and the Semantic Web, as well as in topics related to graph data management.