Scaling Runtimes for Irregular Algorithms to Large-Scale NUMA Systems.
The Galois system can automatically parallelize irregular algorithms written in a serial programming model and execute them efficiently on nonuniform memory access (NUMA) machines. Experimental results for five complex irregular algorithms show that the system scales up to 420× on large NUMA systems...
| Publicado en: | Computer (00189162) Vol. 48; no. 8; pp. 35 - 45 |
|---|---|
| Autores principales: | , |
| Formato: | Artículo |
| Publicado: |
IEEE
Aug2015
|
| Materias: | |
| Acceso en línea: | Ver este registro en EBSCOhost |
| fields | @attributes: recordID: 1 pdfLink: plink: https://search.ebscohost.com/login.aspx?direct=true&db=hlh&AN=108932519&site=ehost-live header: @attributes: shortDbName: hlh uiTerm: 108932519 longDbName: Humanities International Complete uiTag: AN controlInfo: bkinfo: jinfo: jid: 00189162 PUT jtl: Computer (00189162) issn: 00189162 maglogo: N pubinfo: dt: Aug2015 vid: 48 iid: 8 pid: 13605 pub: IEEE artinfo: ui: 108932519 10.1109/MC.2015.229 ppf: 35 ppct: 10 formats: tig: atl: Scaling Runtimes for Irregular Algorithms to Large-Scale NUMA Systems. aug: au: Lenharth, Andrew Pingali, Keshav affil: University of Texas at Austin su: Galois theory Non-uniform memory access Computer storage devices Computer input-output equipment Algorithms sug: subj: Galois theory Non-uniform memory access Computer storage devices Computer input-output equipment Algorithms keyword: ADP amorphous data-parallelism computer architecture Computer graphics Galois Galois fields graph analytics Irregular algorithms irregular applications Large-scale systems memory allocation Memory management nonuniform memory access NUMA parallel programming Runtime Scalability software engineering ab: The Galois system can automatically parallelize irregular algorithms written in a serial programming model and execute them efficiently on nonuniform memory access (NUMA) machines. Experimental results for five complex irregular algorithms show that the system scales up to 420× on large NUMA systems at 512 threads. pubtype: Academic Journal doctype: Article src: R language: English refInfo: copyright: @attributes: flag: Y dt: @attributes: year: 2015 holdings: @attributes: islocal: N |
|---|