describe how we connect to gpu and fetch the information about the hardware
do the performance measurment for different data sizes and differend procesing layout
do the plots for metrics (transfwer data, host to device, device to host, how changes the times depends on data size, repeat for diferent layout.. make a report with tests, to show that you understand that with conclusives)
Measure some ratios (data transfer to procesing time)

Next lab report:
-How a large data structures are handled
-When did it crash
-try different grids
Managed memory