Matches in SemOpenAlex for { <https://semopenalex.org/work/W2557637097> ?p ?o ?g. }
Showing items 1 to 93 of
93
with 100 items per page.
- W2557637097 abstract "High performance dense linear algebra (DLA) libraries often rely on a general matrix multiply (Gemm) kernel that is implemented using assembly or with vector intrinsics. In particular, the real-valued Gemm kernels provide the overwhelming fraction of performance for the complex-valued Gemm kernels, along with the entire level-3 BLAS and many of the real and complex LAPACK routines. Thus,achieving high performance for the Gemm kernel translates into a high performance linear algebra stack above this kernel. However, it is a monumental task for a domain expert to manually implement the kernel for every library-supported architecture. This leads to the belief that the craft of a Gemm kernel is more dark art than science. It is this premise that drives the popularity of autotuning with code generation in the domain of DLA. This paper, instead, focuses on an analytical approach to code generation of the Gemm kernel for different architecture, in order to shed light on the details or voo-doo required for implementing a high performance Gemm kernel. We distill the implementation of the kernel into an even smaller kernel, an outer-product, and analytically determine how available SIMD instructions can be used to compute the outer-product efficiently. We codify this approach into a system to automatically generate a high performance SIMD implementation of the Gemm kernel. Experimental results demonstrate that our approach yields generated kernels with performance that is competitive with kernels implemented manually or using empirical search." @default.
- W2557637097 created "2016-12-08" @default.
- W2557637097 creator A5010052270 @default.
- W2557637097 creator A5011290316 @default.
- W2557637097 creator A5019607600 @default.
- W2557637097 creator A5061722569 @default.
- W2557637097 creator A5062806943 @default.
- W2557637097 date "2016-11-24" @default.
- W2557637097 modified "2023-09-30" @default.
- W2557637097 title "Automating the Last-Mile for High Performance Dense Linear Algebra" @default.
- W2557637097 cites W1964477602 @default.
- W2557637097 cites W1972209410 @default.
- W2557637097 cites W1993445765 @default.
- W2557637097 cites W2002257715 @default.
- W2557637097 cites W2039227678 @default.
- W2557637097 cites W2043275593 @default.
- W2557637097 cites W2064872546 @default.
- W2557637097 cites W2073061372 @default.
- W2557637097 cites W2079658918 @default.
- W2557637097 cites W2084379367 @default.
- W2557637097 cites W2098688018 @default.
- W2557637097 cites W2109341366 @default.
- W2557637097 cites W2114665465 @default.
- W2557637097 cites W2135653967 @default.
- W2557637097 cites W2252007067 @default.
- W2557637097 cites W2291372353 @default.
- W2557637097 cites W2412152731 @default.
- W2557637097 cites W2516525699 @default.
- W2557637097 hasPublicationYear "2016" @default.
- W2557637097 type Work @default.
- W2557637097 sameAs 2557637097 @default.
- W2557637097 citedByCount "2" @default.
- W2557637097 countsByYear W25576370972017 @default.
- W2557637097 countsByYear W25576370972020 @default.
- W2557637097 crossrefType "posted-content" @default.
- W2557637097 hasAuthorship W2557637097A5010052270 @default.
- W2557637097 hasAuthorship W2557637097A5011290316 @default.
- W2557637097 hasAuthorship W2557637097A5019607600 @default.
- W2557637097 hasAuthorship W2557637097A5061722569 @default.
- W2557637097 hasAuthorship W2557637097A5062806943 @default.
- W2557637097 hasConcept C11413529 @default.
- W2557637097 hasConcept C118615104 @default.
- W2557637097 hasConcept C139352143 @default.
- W2557637097 hasConcept C150552126 @default.
- W2557637097 hasConcept C169590947 @default.
- W2557637097 hasConcept C173608175 @default.
- W2557637097 hasConcept C199360897 @default.
- W2557637097 hasConcept C2524010 @default.
- W2557637097 hasConcept C33923547 @default.
- W2557637097 hasConcept C41008148 @default.
- W2557637097 hasConcept C41681595 @default.
- W2557637097 hasConcept C74193536 @default.
- W2557637097 hasConcept C80444323 @default.
- W2557637097 hasConceptScore W2557637097C11413529 @default.
- W2557637097 hasConceptScore W2557637097C118615104 @default.
- W2557637097 hasConceptScore W2557637097C139352143 @default.
- W2557637097 hasConceptScore W2557637097C150552126 @default.
- W2557637097 hasConceptScore W2557637097C169590947 @default.
- W2557637097 hasConceptScore W2557637097C173608175 @default.
- W2557637097 hasConceptScore W2557637097C199360897 @default.
- W2557637097 hasConceptScore W2557637097C2524010 @default.
- W2557637097 hasConceptScore W2557637097C33923547 @default.
- W2557637097 hasConceptScore W2557637097C41008148 @default.
- W2557637097 hasConceptScore W2557637097C41681595 @default.
- W2557637097 hasConceptScore W2557637097C74193536 @default.
- W2557637097 hasConceptScore W2557637097C80444323 @default.
- W2557637097 hasLocation W25576370971 @default.
- W2557637097 hasOpenAccess W2557637097 @default.
- W2557637097 hasPrimaryLocation W25576370971 @default.
- W2557637097 hasRelatedWork W1839773802 @default.
- W2557637097 hasRelatedWork W1972783048 @default.
- W2557637097 hasRelatedWork W1973897068 @default.
- W2557637097 hasRelatedWork W1978642402 @default.
- W2557637097 hasRelatedWork W2114768856 @default.
- W2557637097 hasRelatedWork W2146959317 @default.
- W2557637097 hasRelatedWork W2169492033 @default.
- W2557637097 hasRelatedWork W2222288537 @default.
- W2557637097 hasRelatedWork W2246825238 @default.
- W2557637097 hasRelatedWork W2273809747 @default.
- W2557637097 hasRelatedWork W2617819327 @default.
- W2557637097 hasRelatedWork W2741554038 @default.
- W2557637097 hasRelatedWork W2782176700 @default.
- W2557637097 hasRelatedWork W2886967310 @default.
- W2557637097 hasRelatedWork W2964017942 @default.
- W2557637097 hasRelatedWork W2985842302 @default.
- W2557637097 hasRelatedWork W3036772641 @default.
- W2557637097 hasRelatedWork W3103983137 @default.
- W2557637097 hasRelatedWork W3123057955 @default.
- W2557637097 hasRelatedWork W3205706264 @default.
- W2557637097 isParatext "false" @default.
- W2557637097 isRetracted "false" @default.
- W2557637097 magId "2557637097" @default.
- W2557637097 workType "article" @default.