m2.md

Using ManeFrame II (M2)

A dedicated queue has been setup for Think-Play-Hack, "tph". This queue has 10 nodes, each with 36-cores and 256 GB memory. Two of the nodes also have a NVIDIA P100 GPU with 16 GB of memory. The guest accounts are restricted to the "tph" queue.

Initial Setup

Using Guest Account

1. Request guest account credentials here.
2. Login into M2 via operating system specific instructions

Using SMU Account

1. Login into M2 via operating system specific instructions
2. On M2 run: export MODULEPATH="${MODULEPATH}:/hpc/modules/tph" && module --ignore-cache load python_jupyter r_rstudio

R with RStudio

Using Guest Account

1. On M2 run: srun -p tph -c 1 --mem=6G m2_rstudio
2. Wait for about 30 seconds after the job is allocated and then follow the port forwarding instructions that will be given in the output.

Using SMU Account

1. On M2 run: srun -p tph -A tph -c 1 --mem=6G m2_rstudio All other queues can be used and, if so, remove the -A tph flag.
2. Wait for about 30 seconds after the job is allocated and then follow the port forwarding instructions that will be given in the output.

Python with Jupyter

Using Guest Account

1. On M2 run: srun -p tph -c 1 --mem=6G m2_jupyter_notebook
2. Wait for about 30 seconds after the job is allocated and then follow the port forwarding instructions that will be given in the output.

Using SMU Account

1. On M2 run: srun -p tph -A tph -c 1 --mem=6G m2_jupyter_notebook All other queues can be used and, if so, remove the -A tph flag.
2. Wait for about 30 seconds after the job is allocated and then follow the port forwarding instructions that will be given in the output.