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


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


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

Computer Science > Distributed, Parallel, and Cluster Computing

Title: FT-BLAS: A High Performance BLAS Implementation With Online Fault Tolerance

Abstract: Basic Linear Algebra Subprograms (BLAS) is a core library in scientific computing and machine learning. This paper presents FT-BLAS, a new implementation of BLAS routines that not only tolerates soft errors on the fly, but also provides comparable performance to modern state-of-the-art BLAS libraries on widely-used processors such as Intel Skylake and Cascade Lake. To accommodate the features of BLAS, which contains both memory-bound and computing-bound routines, we propose a hybrid strategy to incorporate fault tolerance into our brand-new BLAS implementation: duplicating computing instructions for memory-bound Level-1 and Level-2 BLAS routines and incorporating an Algorithm-Based Fault Tolerance mechanism for computing-bound Level-3 BLAS routines. Our high performance and low overhead are obtained from delicate assembly-level optimization and a kernel-fusion approach to the computing kernels. Experimental results demonstrate that FT-BLAS offers high reliability and high performance -- faster than Intel MKL, OpenBLAS, and BLIS by up to 3.50%, 22.14% and 21.70%, respectively, for routines spanning all three levels of BLAS we benchmarked, even under hundreds of errors injected per minute.
Comments: camera-ready version at ICS'21: International Conference on Supercomputing 2021 with ISBN updated
Subjects: Distributed, Parallel, and Cluster Computing (cs.DC); Performance (cs.PF)
DOI: 10.1145/3447818.3460364
Cite as: arXiv:2104.00897 [cs.DC]
  (or arXiv:2104.00897v3 [cs.DC] for this version)

Submission history

From: Yujia Zhai [view email]
[v1] Fri, 2 Apr 2021 06:02:58 GMT (3084kb,D)
[v2] Mon, 5 Apr 2021 08:10:07 GMT (3098kb,D)
[v3] Wed, 7 Apr 2021 21:29:55 GMT (3098kb,D)

Link back to: arXiv, form interface, contact.