sujeto_1
Newbie level 1
I'm choosing the path to make my Thesis. Please let me know your opinion about if this how feasible and convenient to do.
I wish to implement 10.000 Cores, each calculating parallel and independently an equation (just as example) like this : A = B^2 + 1 / C where A, B and C are real positive numbers: 1.38, 0.459988,
when all 10.000 cores finish, they communicate with each other in a simple "torus" network to summarize their results; and start again.
Roughly, can you say this makes sense to do it in CUDA? Using a GPU, how difficult would it be to implement? My programming skills are good (assembly, C), but I have never worked with CUDA or GPUs, I'm not sure about the Technology limitations and so.
Finally, do you know a forum more apropiated for CUDA beside NVIDIA website?
I wish to implement 10.000 Cores, each calculating parallel and independently an equation (just as example) like this : A = B^2 + 1 / C where A, B and C are real positive numbers: 1.38, 0.459988,
when all 10.000 cores finish, they communicate with each other in a simple "torus" network to summarize their results; and start again.
Roughly, can you say this makes sense to do it in CUDA? Using a GPU, how difficult would it be to implement? My programming skills are good (assembly, C), but I have never worked with CUDA or GPUs, I'm not sure about the Technology limitations and so.
Finally, do you know a forum more apropiated for CUDA beside NVIDIA website?
