國立虎尾科技大學 |

Designing a Domain-Specific Language for Scalable Development of Large-Scale Computational Science Models.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Designing a Domain-Specific Language for Scalable Development of Large-Scale Computational Science Models./
作者:	Heisler, Eric.
面頁冊數:	1 online resource (105 pages)
附註:	Source: Dissertations Abstracts International, Volume: 85-04, Section: B.
Contained By:	Dissertations Abstracts International85-04B.
標題:	Computer science. -
電子資源:	click for full text (PQDT)
ISBN:	9798380588829

Designing a Domain-Specific Language for Scalable Development of Large-Scale Computational Science Models.
Heisler, Eric.

Designing a Domain-Specific Language for Scalable Development of Large-Scale Computational Science Models. - 1 online resource (105 pages)

Source: Dissertations Abstracts International, Volume: 85-04, Section: B.

Thesis (Ph.D.)--The University of Utah, 2023.

Includes bibliographical references

The numerical solution of differential equations is one of the primary tasks of scientific computing. The mathematical methods have been extensively studied, and the development of corresponding software is a mature field. One direction this has taken is that of domain-specific languages (DSL) that simplify the programming process. This enables domain specialists who are not programming experts to explore computational models using advanced computing techniques.A defining property of a DSL is the abstraction of high-level tasks that hide the low-level programming details. This, unfortunately, results in limited control over algorithm and data-structure design. For research requiring challenging, large-scale computation, these DSLs can limit optimization and usage of sophisticated software tools. This dissertation describes the development of a new DSL that is designed to overcome these limitations. The DSL, named Finch, enables fine-grained control over the resulting code to enable more advanced users to tailor their program to the needs of their problem.Challenging problems in science and engineering can benefit from advanced computational techniques and hardware architectures. Multi-physics and high-dimensional problems like the phonon Boltzmann transport equation require efficient parallel algorithms and data structures to take advantage of hardware resources. Finch provides detailed control over the parallelization and program design. It also enables direct modification of generated code, making this a useful tool for both novice and expert programmers.Challenges also arise during the development of computational models. Exploration of models and numerical techniques can be costly and time-consuming when developed on a large scale or with low-level programming tools. Developing on a coarse-grained version using simple software tools and architectures does not always transfer easily to a full-scale version. Finch addresses this issue through a very flexible and extensible code generation path. The development of models can scale easily from simpler versions to more advanced code without substantial programming effort.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798380588829Subjects--Topical Terms:

573171
Computer science.
Subjects--Index Terms:

Domain-specific languagesIndex Terms--Genre/Form:

554714
Electronic books.

Designing a Domain-Specific Language for Scalable Development of Large-Scale Computational Science Models.
LDR:03542ntm a22003857 4500 001 1143819
005 20240517105001.5
006 m o d
007 cr mn ---uuuuu
008 250605s2023 xx obm 000 0 eng d
020 $a 9798380588829
035 $a (MiAaPQ)AAI30567183
035 $a AAI30567183
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Heisler, Eric. $3 1468618
245 1 0 $a Designing a Domain-Specific Language for Scalable Development of Large-Scale Computational Science Models.
264 0 $c 2023
300 $a 1 online resource (105 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 85-04, Section: B.
500 $a Advisor: Sundar, Hari.
502 $a Thesis (Ph.D.)--The University of Utah, 2023.
504 $a Includes bibliographical references
520 $a The numerical solution of differential equations is one of the primary tasks of scientific computing. The mathematical methods have been extensively studied, and the development of corresponding software is a mature field. One direction this has taken is that of domain-specific languages (DSL) that simplify the programming process. This enables domain specialists who are not programming experts to explore computational models using advanced computing techniques.A defining property of a DSL is the abstraction of high-level tasks that hide the low-level programming details. This, unfortunately, results in limited control over algorithm and data-structure design. For research requiring challenging, large-scale computation, these DSLs can limit optimization and usage of sophisticated software tools. This dissertation describes the development of a new DSL that is designed to overcome these limitations. The DSL, named Finch, enables fine-grained control over the resulting code to enable more advanced users to tailor their program to the needs of their problem.Challenging problems in science and engineering can benefit from advanced computational techniques and hardware architectures. Multi-physics and high-dimensional problems like the phonon Boltzmann transport equation require efficient parallel algorithms and data structures to take advantage of hardware resources. Finch provides detailed control over the parallelization and program design. It also enables direct modification of generated code, making this a useful tool for both novice and expert programmers.Challenges also arise during the development of computational models. Exploration of models and numerical techniques can be costly and time-consuming when developed on a large scale or with low-level programming tools. Developing on a coarse-grained version using simple software tools and architectures does not always transfer easily to a full-scale version. Finch addresses this issue through a very flexible and extensible code generation path. The development of models can scale easily from simpler versions to more advanced code without substantial programming effort.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Computer science. $3 573171
650 4 $a Computer engineering. $3 569006
650 4 $a Information technology. $3 559429
653 $a Domain-specific languages
653 $a Advanced computing
653 $a Data structures
653 $a Low-level programming
655 7 $a Electronic books. $2 local $3 554714
690 $a 0984
690 $a 0489
690 $a 0464
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a The University of Utah. $b School of Computing. $3 1188323
773 0 $t Dissertations Abstracts International $g 85-04B.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30567183 $z click for full text (PQDT)