Matches in SemOpenAlex for { <https://semopenalex.org/work/W3164390940> ?p ?o ?g. }
- W3164390940 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." @default.
- W3164390940 created "2021-06-07" @default.
- W3164390940 creator A5010372747 @default.
- W3164390940 creator A5021674893 @default.
- W3164390940 creator A5035339023 @default.
- W3164390940 creator A5051636078 @default.
- W3164390940 creator A5061737717 @default.
- W3164390940 creator A5076814480 @default.
- W3164390940 date "2021-06-03" @default.
- W3164390940 modified "2023-10-16" @default.
- W3164390940 title "FT-BLAS" @default.
- W3164390940 cites W1541483005 @default.
- W3164390940 cites W1898823215 @default.
- W3164390940 cites W1964588245 @default.
- W3164390940 cites W1986905947 @default.
- W3164390940 cites W1988070283 @default.
- W3164390940 cites W1995746640 @default.
- W3164390940 cites W1997843200 @default.
- W3164390940 cites W2004675401 @default.
- W3164390940 cites W2023856022 @default.
- W3164390940 cites W2030806070 @default.
- W3164390940 cites W2034593585 @default.
- W3164390940 cites W2038238534 @default.
- W3164390940 cites W2046607737 @default.
- W3164390940 cites W2052455844 @default.
- W3164390940 cites W2063639637 @default.
- W3164390940 cites W2064872546 @default.
- W3164390940 cites W2084379367 @default.
- W3164390940 cites W2095928739 @default.
- W3164390940 cites W2102480715 @default.
- W3164390940 cites W2105524676 @default.
- W3164390940 cites W2118832200 @default.
- W3164390940 cites W2125768532 @default.
- W3164390940 cites W2128511938 @default.
- W3164390940 cites W2130076536 @default.
- W3164390940 cites W2130189691 @default.
- W3164390940 cites W2134320686 @default.
- W3164390940 cites W2138692126 @default.
- W3164390940 cites W2150981663 @default.
- W3164390940 cites W2151984682 @default.
- W3164390940 cites W2156514327 @default.
- W3164390940 cites W2169596872 @default.
- W3164390940 cites W2170196949 @default.
- W3164390940 cites W2229245554 @default.
- W3164390940 cites W2252007067 @default.
- W3164390940 cites W2292469857 @default.
- W3164390940 cites W2296204683 @default.
- W3164390940 cites W2343351966 @default.
- W3164390940 cites W2411755313 @default.
- W3164390940 cites W2412349256 @default.
- W3164390940 cites W2418331349 @default.
- W3164390940 cites W2419664341 @default.
- W3164390940 cites W2485331474 @default.
- W3164390940 cites W2621842477 @default.
- W3164390940 cites W2647773517 @default.
- W3164390940 cites W2767260595 @default.
- W3164390940 cites W2767321582 @default.
- W3164390940 cites W2767694495 @default.
- W3164390940 cites W2949981335 @default.
- W3164390940 cites W2986161099 @default.
- W3164390940 cites W3105862567 @default.
- W3164390940 cites W4229666556 @default.
- W3164390940 doi "https://doi.org/10.1145/3447818.3460364" @default.
- W3164390940 hasPublicationYear "2021" @default.
- W3164390940 type Work @default.
- W3164390940 sameAs 3164390940 @default.
- W3164390940 citedByCount "3" @default.
- W3164390940 countsByYear W31643909402022 @default.
- W3164390940 countsByYear W31643909402023 @default.
- W3164390940 crossrefType "proceedings-article" @default.
- W3164390940 hasAuthorship W3164390940A5010372747 @default.
- W3164390940 hasAuthorship W3164390940A5021674893 @default.
- W3164390940 hasAuthorship W3164390940A5035339023 @default.
- W3164390940 hasAuthorship W3164390940A5051636078 @default.
- W3164390940 hasAuthorship W3164390940A5061737717 @default.
- W3164390940 hasAuthorship W3164390940A5076814480 @default.
- W3164390940 hasBestOaLocation W31643909401 @default.
- W3164390940 hasConcept C111919701 @default.
- W3164390940 hasConcept C114614502 @default.
- W3164390940 hasConcept C173608175 @default.
- W3164390940 hasConcept C2779960059 @default.
- W3164390940 hasConcept C33923547 @default.
- W3164390940 hasConcept C41008148 @default.
- W3164390940 hasConcept C63540848 @default.
- W3164390940 hasConcept C74193536 @default.
- W3164390940 hasConcept C83283714 @default.
- W3164390940 hasConceptScore W3164390940C111919701 @default.
- W3164390940 hasConceptScore W3164390940C114614502 @default.
- W3164390940 hasConceptScore W3164390940C173608175 @default.
- W3164390940 hasConceptScore W3164390940C2779960059 @default.
- W3164390940 hasConceptScore W3164390940C33923547 @default.
- W3164390940 hasConceptScore W3164390940C41008148 @default.
- W3164390940 hasConceptScore W3164390940C63540848 @default.
- W3164390940 hasConceptScore W3164390940C74193536 @default.
- W3164390940 hasConceptScore W3164390940C83283714 @default.
- W3164390940 hasLocation W31643909401 @default.
- W3164390940 hasLocation W31643909402 @default.
- W3164390940 hasLocation W31643909403 @default.
- W3164390940 hasOpenAccess W3164390940 @default.
- W3164390940 hasPrimaryLocation W31643909401 @default.