We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.DC

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Information Theory

Title: Straggler Mitigation through Unequal Error Protection for Distributed Matrix Multiplication

Abstract: Large-scale machine learning and data mining methods routinely distribute computations across multiple agents to parallelize processing. The time required for computation at the agents is affected by the availability of local resources giving rise to the "straggler problem" in which the computation results are held back by unresponsive agents. For this problem, linear coding of the matrix sub-blocks can be used to introduce resilience toward straggling. The Parameter Server (PS) utilizes a channel code and distributes the matrices to the workers for multiplication. It then produces an approximation to the desired matrix multiplication using the results of the computations received at a given deadline. In this paper, we propose to employ Unequal Error Protection (UEP) codes to alleviate the straggler problem. The resiliency level of each sub-block is chosen according to its norm as blocks with larger norms have higher effects on the result of the matrix multiplication. We validate the effectiveness of our scheme both theoretically and through numerical evaluations. We derive a theoretical characterization of the performance of UEP using random linear codes, and compare it the case of equal error protection. We also apply the proposed coding strategy to the computation of the back-propagation step in the training of a Deep Neural Network (DNN), for which we investigate the fundamental trade-off between precision and the time required for the computations.
Comments: 6 pages, 6 figures
Subjects: Information Theory (cs.IT); Distributed, Parallel, and Cluster Computing (cs.DC)
Cite as: arXiv:2011.02749 [cs.IT]
  (or arXiv:2011.02749v2 [cs.IT] for this version)

Submission history

From: Busra Tegin [view email]
[v1] Thu, 5 Nov 2020 10:43:32 GMT (36kb)
[v2] Fri, 19 Mar 2021 08:24:36 GMT (166kb)

Link back to: arXiv, form interface, contact.