Hello everyone,
I’m working on a simulation at a time, a path planning, which partly runs on jCUDA to calculate the path but the environment runs on the CPU using jME 2. And I got the idea to create a framework for us to make use of a cluster of GPU’s we have in order to generate the simulation in a Cloud Computing environment, but I have the idea to put the simulation to run as a web application, I noticed that there is a way to do
it with jME.
I would like to ask for help from more experienced to criticize my idea because I do not know if this is feasible, leave part of the processing on the server (Cluster) and visualization on the client, as in Massively Multiplayer Online Game (MMOG).
I thank the anonymous reviewers for their constructive and helpful comments.
It shouldn’t really be a problem to do this, though it might be interesting to see about running the simulation on the client hardware as well (checking of course if the current graphics hardware supports CUDA)… You should be able to pack up the jCUDA distribution with the rest of your jar’s and native libraries.
1 Like
Actually I want to run the server part of CUDA as a Cloud Application …
Aye, then what you’re striving for is really pretty simple and within the expected use cases of jME 
As a Web Start application or an Applet you can do just about anything that a free-standing Java application can do.