Search for: All records

We describe a paradigm for designing parallel algorithms via approximation, and illustrate it on the b-edgecover problem. A b-edgecover of minimum weight in a graph is a subset $$C$$ of its edges such that at least a specified number $b(v)$ of edges in $$C$$ is incident on each vertex $$v$$, and the sum of the edge weights in $$C$$ is minimum. The Greedy algorithm and a variant, the LSE algorithm, provide $3/2$-approximation guarantees in the worst-case for this problem, but these algorithms have limited parallelism. Hence we design two new $$2$$-approximation algorithms with greater concurrency. The MCE algorithm reduces the computation of a b-edgecover to that of finding a b'-matching, by exploiting the relationship between these subgraphs in an approximation context. The LSE-NW is derived from the LSEalgorithm using static edge weights rather than dynamically computing effective edge weights. This relaxation gives LSE a worse approximation guarantee but makes it more amenable to parallelization. We prove that both the MCE and LSE-NW algorithms compute the same b-edgecover with at most twice the weight of the minimum weight edge cover. In practice, the $$2$$-approximation and $3/2$-approximation algorithms compute edge covers of weight within $$10\%$$ the optimal. We implement three of the approximation algorithms, MCE, LSE, and LSE-NW on shared memory multi-core machines, including an Intel Xeon and an IBM Power8 machine with 8 TB memory. The MCE algorithm is the fastest of these by an order of magnitude or more. It computes an edge cover in a graph with billions of edges in $20$ seconds using two hundred threads on the IBM Power8. We also show that the parallel depth and work can be bounded for the Suitor and b-Suitor algorithms when edge weights are random. 

We describe a 3/2-approximation algorithm, \lse, for computing a b-edgecover of minimum weight in a graph with weights on the edges. The b-edgecover problem is a generalization of the better-known Edge Cover problem in graphs, where the objective is to choose a subset C of edges in the graph such that at least a specified number b(v) of edges in C are incident on each vertex v. In the weighted b-edgecover problem, we minimize the sum of the weights of the edges in C. We prove that the Locally Subdominant edge (LSE) algorithm computes the same b-edge cover as the one obtained by the Greedy algorithm for the problem. However, the Greedy algorithm requires edges to be sorted by their effective weights, and these weights need to be updated after each iteration. These requirements make the Greedy algorithm sequential and impractical for massive graphs. The LSE algorithm avoids the sorting step, and is amenable for parallelization. We implement the algorithm on a serial machine and compare its performance against a collection of approximation algorithms for the b-edge cover problem. Our results show that the algorithm is 3 to 5 times faster than the Greedy algorithm on a serial processor. The approximate edge covers obtained by the LSE algorithm have weights greater by at most 17% of the optimal weight for problems where we could compute the latter. We also investigate the relationship between the b-edge cover and the b-matching problems, show that the latter has a faster implementation since edge weights are static in this algorithm, and obtain a heuristic solution for the former from the latter. 

A b-matching is a subset of edges M such that at most b(v) edges in M are incident on each vertex v, where b(v) is specified. We present a distributed-memory parallel algorithm, \bsuitor, that computes a b-matching with more than half the maximum weight in a graph with weights on the edges. The approximation algorithm is designed to have high concurrency and low time complexity. We organize the implementation of the algorithm in terms of asynchronous super-steps that combine computation and communication, and balance the computational work and frequency of communication to obtain high performance. Since the performance of the b-suitor algorithm is strongly influenced by communication, we present several strategies to reduce the communication volume. We implement the algorithm using a hybrid strategy where inter-node communication uses MPI and intra-node computation is done with OpenMP threads. We demonstrate strong and weak scaling of b-suitor up to 16,000 processors on two supercomputers at NERSC. We compute a b-matching in a graph with 2 billion edges in under 4 seconds using 16,000 processors.