I am not sure if my question fits in this category, but it seems to me to be the closest to what I would like to ask.
My wish is to buy a gaming laptop (such as Asus ROG, Dell G3, some Alienware, … or what have you on the market) with Nvidia GPUs, install Linux (my favorite flavor is Ubuntu), Nvidia compilers and, there I go, have a machine which I can use to port my existing codes for fluid flow simulation, written either in C++ or Fortran 2003 accelerated with MPI only for the time being, to GPUs.
Now, my question would be: is it possible at all? I see a number of potential problems here, such as:
Will it be possible to install Linux on any gaming laptop? (I never had a gaming nor do I play games anymore, but I presume they come with Windows pre-installed, not really an environment I find friendly for code development.)
Assuming that the installation of Linux on a gaming laptop is not an issue. Will installing Nvidia HPC-SDK be sufficient to start developing GPU-accelerated programs with OpenACC? I mean, can one expect issues with compilers not being able to recognize Nvidia GPUs or launching simulations on them? Will it be possible to use the GPU card for programming, or will the system believe it is just for display?
If neither of the above is an issue, which graphical card to chose? Judging from the kind of work I am doing (MPI) I would certainly need a multi-core multi-threaded CPU and a GPU with decent amount of memory, but have no clue which one to chose? For example, Dell G3 comes with Nvidia GTX 1660, but Asus ROG Zephyrus G14 comes with Nvidia RTX 2060 Max-Q. Another example, Dell Alienware https://www.dell.com/en-us/shop/cty/pdp/spd/alienware-m15-r2-laptop : i9 CPU, 16GB + Nvidia GeForce RTX 2080 8GB looks like an attractive combination to me, is it?
That would essentially be what I wanted to ask. Any advice would be highly appreciated.