Computer Science - Groups

Algorithm Engineering

Ulrich Meyer

Meyer's work focuses on the design and engineering of algorithms and
data structures for very large data sets, especially in the context of
parallelism (Shared-Memory, clusters, GPUs, ...), memory hierarchies
(caches, SSDs, hard disks, ...), and energy efficiency. The spectrum
ranges from theoretical worst-case results and average-case analyses to
experimental evaluation of heuristics.

Large amounts of data are often represented in the form of graphs. The
structured traversal of these graphs is a notoriously difficult problem
both in the parallel environment and for memory hierarchies. Using
algorithms with better access patterns in worst-case efficient
implementations, we have already achieved more than a thousand-fold
speed-up over the state-of-the-art. Similarly, we developed DFS Search
heuristics that can be used for memory-efficient analysis of web graphs,
telecommunication data (call graphs) or large social networks. Recent
work has focused on memory-efficient graph exploration in dynamically
changing data, approximate solutions, as well as the efficient
generation of randomized graph test data.

Fellow Detail

Selected Funding:

Logo SPP 1736 Algorithms for Big Data

The Algorithm Engineering group coordinates the DFG Priority Programme SPP 1736 "Algorithms for Big Data"