Matches in SemOpenAlex for { <https://semopenalex.org/work/W4231267341> ?p ?o ?g. }
Showing items 1 to 95 of
95
with 100 items per page.
- W4231267341 endingPage "638" @default.
- W4231267341 startingPage "623" @default.
- W4231267341 abstract "A Fused Multiply-Add (FMA) instruction is currently available in many general-purpose processors. It increases performance by reducing latency of dependent operations and increases precision by computing the result as an indivisible operation with no intermediate rounding. However, since the arithmetic behavior of a single-rounding FMA operation is different than independent FP multiply followed by FP add instructions, some algorithms require significant revalidation and rewriting efforts to work as expected when they are compiled to operate with FMA--a cost that developers may not be willing to pay. Because of that, abundant legacy applications are not able to utilize FMA instructions. In this paper we propose a novel HW/SW collaborative technique that is able to efficiently execute workloads with increased utilization of FMA, by adding the option to get the same numerical result as separate FP multiply and FP add pairs. In particular, we extended the host ISA of a HW/SW co-designed processor with a new Combined Multiply-Add (CMA) instruction that performs an FMA operation with an intermediate rounding. This new instruction is used by a transparent dynamic translation software layer that uses a speculative instruction-fusion optimization to transform FP multiply and FP add sequences into CMA instructions. The FMA unit has been slightly modified to support both single-rounding and double-rounding fused instructions without increasing their latency and to provide a conservative fall-back path in case of mispeculation. Evaluation on a cycle-accurate timing simulator showed that CMA improved SPECfp performance by 6.3% and reduced executed instructions by 4.7%." @default.
- W4231267341 created "2022-05-12" @default.
- W4231267341 creator A5000686736 @default.
- W4231267341 creator A5001719897 @default.
- W4231267341 creator A5015144284 @default.
- W4231267341 creator A5015318368 @default.
- W4231267341 creator A5026926090 @default.
- W4231267341 creator A5046385142 @default.
- W4231267341 creator A5080905072 @default.
- W4231267341 date "2014-02-24" @default.
- W4231267341 modified "2023-09-26" @default.
- W4231267341 title "Speculative hardware/software co-designed floating-point multiply-add fusion" @default.
- W4231267341 cites W1979072566 @default.
- W4231267341 cites W1980480935 @default.
- W4231267341 cites W1983778743 @default.
- W4231267341 cites W2036482573 @default.
- W4231267341 cites W2057234250 @default.
- W4231267341 cites W2057543661 @default.
- W4231267341 cites W2070273314 @default.
- W4231267341 cites W2088020108 @default.
- W4231267341 cites W2100974190 @default.
- W4231267341 cites W2109611156 @default.
- W4231267341 cites W2120371291 @default.
- W4231267341 cites W2121764011 @default.
- W4231267341 cites W2130123999 @default.
- W4231267341 cites W2134663450 @default.
- W4231267341 cites W2140351961 @default.
- W4231267341 cites W2148865465 @default.
- W4231267341 cites W2150023094 @default.
- W4231267341 cites W2153456949 @default.
- W4231267341 cites W2168020750 @default.
- W4231267341 cites W3141428879 @default.
- W4231267341 cites W4236971743 @default.
- W4231267341 cites W4251914687 @default.
- W4231267341 doi "https://doi.org/10.1145/2654822.2541978" @default.
- W4231267341 hasPublicationYear "2014" @default.
- W4231267341 type Work @default.
- W4231267341 citedByCount "0" @default.
- W4231267341 crossrefType "journal-article" @default.
- W4231267341 hasAuthorship W4231267341A5000686736 @default.
- W4231267341 hasAuthorship W4231267341A5001719897 @default.
- W4231267341 hasAuthorship W4231267341A5015144284 @default.
- W4231267341 hasAuthorship W4231267341A5015318368 @default.
- W4231267341 hasAuthorship W4231267341A5026926090 @default.
- W4231267341 hasAuthorship W4231267341A5046385142 @default.
- W4231267341 hasAuthorship W4231267341A5080905072 @default.
- W4231267341 hasBestOaLocation W42312673411 @default.
- W4231267341 hasConcept C110305270 @default.
- W4231267341 hasConcept C111919701 @default.
- W4231267341 hasConcept C136625980 @default.
- W4231267341 hasConcept C156972235 @default.
- W4231267341 hasConcept C173608175 @default.
- W4231267341 hasConcept C2777904410 @default.
- W4231267341 hasConcept C33923547 @default.
- W4231267341 hasConcept C41008148 @default.
- W4231267341 hasConcept C49154492 @default.
- W4231267341 hasConcept C76155785 @default.
- W4231267341 hasConcept C82876162 @default.
- W4231267341 hasConcept C84211073 @default.
- W4231267341 hasConcept C9390403 @default.
- W4231267341 hasConcept C94375191 @default.
- W4231267341 hasConceptScore W4231267341C110305270 @default.
- W4231267341 hasConceptScore W4231267341C111919701 @default.
- W4231267341 hasConceptScore W4231267341C136625980 @default.
- W4231267341 hasConceptScore W4231267341C156972235 @default.
- W4231267341 hasConceptScore W4231267341C173608175 @default.
- W4231267341 hasConceptScore W4231267341C2777904410 @default.
- W4231267341 hasConceptScore W4231267341C33923547 @default.
- W4231267341 hasConceptScore W4231267341C41008148 @default.
- W4231267341 hasConceptScore W4231267341C49154492 @default.
- W4231267341 hasConceptScore W4231267341C76155785 @default.
- W4231267341 hasConceptScore W4231267341C82876162 @default.
- W4231267341 hasConceptScore W4231267341C84211073 @default.
- W4231267341 hasConceptScore W4231267341C9390403 @default.
- W4231267341 hasConceptScore W4231267341C94375191 @default.
- W4231267341 hasIssue "1" @default.
- W4231267341 hasLocation W42312673411 @default.
- W4231267341 hasOpenAccess W4231267341 @default.
- W4231267341 hasPrimaryLocation W42312673411 @default.
- W4231267341 hasRelatedWork W1520019936 @default.
- W4231267341 hasRelatedWork W2014186618 @default.
- W4231267341 hasRelatedWork W2027965104 @default.
- W4231267341 hasRelatedWork W2064131220 @default.
- W4231267341 hasRelatedWork W2073750546 @default.
- W4231267341 hasRelatedWork W2158464203 @default.
- W4231267341 hasRelatedWork W2183609869 @default.
- W4231267341 hasRelatedWork W2470712352 @default.
- W4231267341 hasRelatedWork W2540610482 @default.
- W4231267341 hasRelatedWork W4376144198 @default.
- W4231267341 hasVolume "42" @default.
- W4231267341 isParatext "false" @default.
- W4231267341 isRetracted "false" @default.
- W4231267341 workType "article" @default.