thesis update
for the past few weeks i have been preoccupied by my study on a projective load balancing algorithm as well as a parallel multiple population genetic algorithm with species migration solution to the assymetric travelling salesman problem. i have run a pretest on the 16 node beowulf cluster i have setup at the Institute of Computer Science at the University of the Philippines at Los Banos (UPLB ICS) by getting the area under the curve f(x) = 1/x^3 + 1/x^2 + 1/x from 1 to 1000. and right now, i have very encouraging findings regarding the performance speedup of the 16 node parallel computation compared to a sequqntial implementation of the solution. however, the findings will be posted in due time.
right now i am writing a couple of papers regarding 1) the computational speedup of parallel computation using clusters and try and a mathematical model on the relation among the number of processors, the amount of computational time it takes to complete the run of a solution, and the size of the problem/solution at hand implemented in parallel and 2) a document on the formal steps and processes in setting up beowulf clusters according to my analysis and experience.
i will also be starting on a couple of papers i would like to submit to the 4th Philippine Computing Science Congress, one being on the parallel multiple population genetic algorithm with species migration solution to the assymetric travelling salesman problem, and another on the findings regarding my thesis on projective load balancing for heterogenous paralllel computing clusters.
now that they're out of the way, i should be on my way to working on them. now i've put them online, and in perspective which compels me to actually start and do something.
still chillin'...
for the past few weeks i have been preoccupied by my study on a projective load balancing algorithm as well as a parallel multiple population genetic algorithm with species migration solution to the assymetric travelling salesman problem. i have run a pretest on the 16 node beowulf cluster i have setup at the Institute of Computer Science at the University of the Philippines at Los Banos (UPLB ICS) by getting the area under the curve f(x) = 1/x^3 + 1/x^2 + 1/x from 1 to 1000. and right now, i have very encouraging findings regarding the performance speedup of the 16 node parallel computation compared to a sequqntial implementation of the solution. however, the findings will be posted in due time.
right now i am writing a couple of papers regarding 1) the computational speedup of parallel computation using clusters and try and a mathematical model on the relation among the number of processors, the amount of computational time it takes to complete the run of a solution, and the size of the problem/solution at hand implemented in parallel and 2) a document on the formal steps and processes in setting up beowulf clusters according to my analysis and experience.
i will also be starting on a couple of papers i would like to submit to the 4th Philippine Computing Science Congress, one being on the parallel multiple population genetic algorithm with species migration solution to the assymetric travelling salesman problem, and another on the findings regarding my thesis on projective load balancing for heterogenous paralllel computing clusters.
now that they're out of the way, i should be on my way to working on them. now i've put them online, and in perspective which compels me to actually start and do something.
still chillin'...
Comments
Post a Comment