Problem with large data of Google trace in simulation running

[duaa h]

My device has 2 GPUs linked via SLI
How can I make the performance of cloudsim especially I am trying to run Google trace and facing difficulty in running large data even for data of one day.
My questions are:
1-Is there a way to install cloudsim and benefit from the linked GPUs? Is there a document that explain how?
2-or should I try to run cloudsim and the data should be divided by benefiting from the GPUs ? How cloud I do so?

[Manoel Campos]

How can I make the performance of cloudsim especially I am trying to run Google trace and facing difficulty in running large data even for data of one day.

Google cluster traces are large files that impose a large memory consumption and CPU overhead. 

 

1-Is there a way to install cloudsim and benefit from the linked GPUs? Is there a document that explain how? 

2-or should I try to run cloudsim and the data should be divided by benefiting from the GPUs ? How cloud I do so?

If you mean CloudSim Plus, I've never tried that and it's out of the project's scope.

Possibly there are some way to run Java apps in the GPU.