Matches in SemOpenAlex for { <https://semopenalex.org/work/W2018421763> ?p ?o ?g. }
Showing items 1 to 87 of
87
with 100 items per page.
- W2018421763 abstract "In this paper we explore the capability and flexibility of FPGA solutions in a sense to accelerate scientific computing applications which require very high precision arithmetic, based on 128-bit or even 256-bit floating-point number representation.This paper addresses the accuracy when performing LU decomposition on large-scale matrices. In future ExaScale computing environments, accuracy errors are expected to increase up to a level which leaves only 11 significant bits in the mantissa. This is caused by the required large amount of accumulation operations which are in the order of O(n3). Using exact long fixed-point numbers instead of usual floatingpoint numbers in the accumulation process, leads to exact accumulation results with only one bit error, originated by the rounding in the last normalization step. We have developed two types of High Precision Multiplication and Accumulation (HP-MAC), for Double-Double (128 bits) and Quad-Double (256 bits) floating-point, respectively, and implemented them into FPGA devices. We propose a two-level RAM banks scheme to store and add long fixed-point numbers with minimized crucial data paths lengths. We also introduce a scheme of partial summation to enhance the pipeline throughput of MAC operations, by dividing the summation function into 4 partial operations, processed in 4 banks. To prove the concept, we prototyped six 128-bit HP-MAC units into a Xilinx Virtex-5 XC5VLX330 FPGA chip and performed LU decomposition.The experimental results show accuracy improvement of 10 to 24 bits, compared to a software approach with similar precision arithmetic. Moreover, our LU decomposition implementation, based on FPGA running at 133MHz, achieves 29X--56X better performance and much lower power consumption compared to the use of a software-based library running on an Intel Core2 Quad Q8200 CPU at 2.33GHz." @default.
- W2018421763 created "2016-06-24" @default.
- W2018421763 creator A5007279267 @default.
- W2018421763 creator A5026041659 @default.
- W2018421763 creator A5035113789 @default.
- W2018421763 creator A5051680867 @default.
- W2018421763 creator A5055646953 @default.
- W2018421763 creator A5057354933 @default.
- W2018421763 date "2010-06-02" @default.
- W2018421763 modified "2023-10-16" @default.
- W2018421763 title "FPGA accelerating double/quad-double high precision floating-point applications for ExaScale computing" @default.
- W2018421763 cites W1506162491 @default.
- W2018421763 cites W1969213662 @default.
- W2018421763 cites W1975602595 @default.
- W2018421763 cites W2002064155 @default.
- W2018421763 cites W2053987390 @default.
- W2018421763 cites W2061624656 @default.
- W2018421763 cites W2104281151 @default.
- W2018421763 cites W2106191211 @default.
- W2018421763 cites W2107996047 @default.
- W2018421763 cites W2127669225 @default.
- W2018421763 cites W2132367502 @default.
- W2018421763 cites W2132633136 @default.
- W2018421763 cites W2165869826 @default.
- W2018421763 doi "https://doi.org/10.1145/1810085.1810129" @default.
- W2018421763 hasPublicationYear "2010" @default.
- W2018421763 type Work @default.
- W2018421763 sameAs 2018421763 @default.
- W2018421763 citedByCount "10" @default.
- W2018421763 countsByYear W20184217632012 @default.
- W2018421763 countsByYear W20184217632013 @default.
- W2018421763 countsByYear W20184217632014 @default.
- W2018421763 countsByYear W20184217632015 @default.
- W2018421763 countsByYear W20184217632016 @default.
- W2018421763 countsByYear W20184217632017 @default.
- W2018421763 crossrefType "proceedings-article" @default.
- W2018421763 hasAuthorship W2018421763A5007279267 @default.
- W2018421763 hasAuthorship W2018421763A5026041659 @default.
- W2018421763 hasAuthorship W2018421763A5035113789 @default.
- W2018421763 hasAuthorship W2018421763A5051680867 @default.
- W2018421763 hasAuthorship W2018421763A5055646953 @default.
- W2018421763 hasAuthorship W2018421763A5057354933 @default.
- W2018421763 hasConcept C111919701 @default.
- W2018421763 hasConcept C149635348 @default.
- W2018421763 hasConcept C173608175 @default.
- W2018421763 hasConcept C2524010 @default.
- W2018421763 hasConcept C2778837361 @default.
- W2018421763 hasConcept C28719098 @default.
- W2018421763 hasConcept C33923547 @default.
- W2018421763 hasConcept C35912277 @default.
- W2018421763 hasConcept C41008148 @default.
- W2018421763 hasConcept C42935608 @default.
- W2018421763 hasConcept C83283714 @default.
- W2018421763 hasConcept C84211073 @default.
- W2018421763 hasConcept C9390403 @default.
- W2018421763 hasConceptScore W2018421763C111919701 @default.
- W2018421763 hasConceptScore W2018421763C149635348 @default.
- W2018421763 hasConceptScore W2018421763C173608175 @default.
- W2018421763 hasConceptScore W2018421763C2524010 @default.
- W2018421763 hasConceptScore W2018421763C2778837361 @default.
- W2018421763 hasConceptScore W2018421763C28719098 @default.
- W2018421763 hasConceptScore W2018421763C33923547 @default.
- W2018421763 hasConceptScore W2018421763C35912277 @default.
- W2018421763 hasConceptScore W2018421763C41008148 @default.
- W2018421763 hasConceptScore W2018421763C42935608 @default.
- W2018421763 hasConceptScore W2018421763C83283714 @default.
- W2018421763 hasConceptScore W2018421763C84211073 @default.
- W2018421763 hasConceptScore W2018421763C9390403 @default.
- W2018421763 hasFunder F4320321001 @default.
- W2018421763 hasFunder F4320321540 @default.
- W2018421763 hasLocation W20184217631 @default.
- W2018421763 hasOpenAccess W2018421763 @default.
- W2018421763 hasPrimaryLocation W20184217631 @default.
- W2018421763 hasRelatedWork W2022020111 @default.
- W2018421763 hasRelatedWork W2083682931 @default.
- W2018421763 hasRelatedWork W2134728271 @default.
- W2018421763 hasRelatedWork W2257153718 @default.
- W2018421763 hasRelatedWork W2366969118 @default.
- W2018421763 hasRelatedWork W2392910503 @default.
- W2018421763 hasRelatedWork W2400010384 @default.
- W2018421763 hasRelatedWork W3011295035 @default.
- W2018421763 hasRelatedWork W3081870748 @default.
- W2018421763 hasRelatedWork W2083974768 @default.
- W2018421763 isParatext "false" @default.
- W2018421763 isRetracted "false" @default.
- W2018421763 magId "2018421763" @default.
- W2018421763 workType "article" @default.