Rechercher
Derniers sujets
SETI at home or out of home.. at work:)! Lol?
Page 1 sur 1
SETI at home or out of home.. at work:)! Lol?
Due my 'invest.' since sometimes in BOINC, and in SETI, i promote some of their activities in search of alien's electronics 'signals', related among others places at :
http://www.cs.ucdavis.edu/
--
http://www.cs.ucdavis.edu/
--
Dernière édition par Milux le Mer 19 Jan - 21:01, édité 1 fois
Milux- Messages : 2794
Date d'inscription : 30/08/2010
Re: SETI at home or out of home.. at work:)! Lol?
Resume came as :
'Project description:
Find an item on the Web--a research paper, article or possibly even a slide presentation--that reports on the use of CUDA, MPI or OpenMP to a real problem in some field of application, such as medicine, business, genomics, social networks and so on. Your report will consist of an analysis of this item.
The real spirit of this project is to immerse yourself in a field of application of computer science. An excellent example is GPU Acceleration of Cutoff Pair Potentials for Molecular Modeling Applications, by Rodrigues et al, available from UCD IP addresses here. (I found this paper by plugging "CUDA thread divergence" into Google Scholar.)
What make the above paper so suitable is that it does explain the application, in this case molecular modeling. In addition, it goes through the equations that arise in that application, stating what they mean, what their significance is, and why parallel processing is needed.
By contrast, the paper Biomedical Image Analysis on a Cooperative Cluster of GPUs and Multicores here, by Hartley et al would be a poor choice. Its description of the application is perfunctory at best, just a side remark, really. There is no real description given of the math of the problem to be solved.
On the other hand, if a group were to do their Final Report on that Hartley paper, it would be fine if they supplemented it with other materials, say from references in the paper's bibliography.
Reporting on an article which "applies CS to CS," in which CUDA code is written to produce a fast sorting algorithm, for instance, is strongly discouraged. If you feel strongly about some article of this type in particular, I will consider it, but please try to find something that fits the spirit of the project first.
Your report should address the following questions:
What was the nature of the application?
Why was parallel computation needed?
What problems did they anticipate/encounter, and how did they solve them?
What recommendations would you have to further increase performance?
Suppose they had used one of the other two languages, not the one they used. Give very brief pseudocode outlines of how those other languages might be used, and explain what issues would likely arise:)!'
--
'Project description:
Find an item on the Web--a research paper, article or possibly even a slide presentation--that reports on the use of CUDA, MPI or OpenMP to a real problem in some field of application, such as medicine, business, genomics, social networks and so on. Your report will consist of an analysis of this item.
The real spirit of this project is to immerse yourself in a field of application of computer science. An excellent example is GPU Acceleration of Cutoff Pair Potentials for Molecular Modeling Applications, by Rodrigues et al, available from UCD IP addresses here. (I found this paper by plugging "CUDA thread divergence" into Google Scholar.)
What make the above paper so suitable is that it does explain the application, in this case molecular modeling. In addition, it goes through the equations that arise in that application, stating what they mean, what their significance is, and why parallel processing is needed.
By contrast, the paper Biomedical Image Analysis on a Cooperative Cluster of GPUs and Multicores here, by Hartley et al would be a poor choice. Its description of the application is perfunctory at best, just a side remark, really. There is no real description given of the math of the problem to be solved.
On the other hand, if a group were to do their Final Report on that Hartley paper, it would be fine if they supplemented it with other materials, say from references in the paper's bibliography.
Reporting on an article which "applies CS to CS," in which CUDA code is written to produce a fast sorting algorithm, for instance, is strongly discouraged. If you feel strongly about some article of this type in particular, I will consider it, but please try to find something that fits the spirit of the project first.
Your report should address the following questions:
What was the nature of the application?
Why was parallel computation needed?
What problems did they anticipate/encounter, and how did they solve them?
What recommendations would you have to further increase performance?
Suppose they had used one of the other two languages, not the one they used. Give very brief pseudocode outlines of how those other languages might be used, and explain what issues would likely arise:)!'
--
Milux- Messages : 2794
Date d'inscription : 30/08/2010
Re: SETI at home or out of home.. at work:)! Lol?
4 else everything is there....., need to be upload 4 that:)!
http://www.cs.ucdavis.edu/department/news/Becker_Peri_Final_Report.pdf
--
http://www.cs.ucdavis.edu/department/news/Becker_Peri_Final_Report.pdf
--
Milux- Messages : 2794
Date d'inscription : 30/08/2010
Page 1 sur 1
Permission de ce forum:
Vous ne pouvez pas répondre aux sujets dans ce forum
Mar 19 Mai - 19:29 par Milux
» News #Open source, #FOSS , #Linux et autres #Freewares et #Free-mobile
Jeu 30 Mai - 15:29 par Milux
» Quelques news 'Portnawak' & autres bilevesées:(!
Jeu 30 Mai - 15:12 par Milux
» TinyLinux : Pupy
Mar 28 Mai - 17:41 par Milux
» Topics of interrest, FOSS et évolutions & relayés sur Sccop.it :)!
Jeu 16 Mai - 13:42 par Milux
» Après F.B. google Buzz & iGoogle, G+1 arrive:>?
Jeu 16 Mai - 11:41 par Milux
» Buzz, Blogs, FB et pages de sites censurées - Google +:<? -
Jeu 16 Mai - 11:33 par Milux
» Jaime pas, mais j'en parle aussi :)?
Mer 1 Mai - 18:30 par Milux
» J'aime, donc j'en parle:)!
Mer 17 Avr - 15:24 par Milux