The teaching GPU cluster consists of a head node and three GPU compute nodes. The head node is a CPU-only machine while each compute node has multiple GPUs. The cluster is controlled by the Slurm workload manager. To use the compute nodes you must submit jobs via the head node, namely tsubmit.math.private.uwaterloo.ca. There is one general-use GPU node (gpu-pt1-01) and two course-specific GPU nodes (gpu-pt1-02 and gpu-pt1-03).
-
Head
node
Node name slurm-pt2.math.private.uwaterloo.ca,
alias tsubmit.math.private.uwaterloo.ca#Nodes 1 CPU model (2) Intel(R) Xeon(R) Gold 6150 CPU @ 2.70GHz #Cores/Node 36 Threads per core 2 System Memory/Node 188 GB -
General-use
GPU
compute
node
This
GPU
machine
is
not
restricted
to
particular
courses,
so
any
undergraduate
student
registered
in
a
non-CS
Math
course
may
use
it.
(CS
students
should
use
CS
resources.)
Node names gpu-pt1-01.math.private.uwaterloo.ca #Nodes 1 CPU model (2 per node) Intel(R) Xeon(R) CPU E5-2680 v4 @ 2.40GHz #Cores/Node 28 Threads per core 2 System Memory/Node 128 GB GPU Type NVIDIA Tesla K80 # GPU devices/Node 8 GPU memory per device 12 GB -
Course-specific
GPU
compute
nodes
These
machines
are
dedicated
to certain
courses
that
involve
GPU
computing,
such
as
machine
learning. Only
students
registered
in
the
GPU-related
courses
may
use
these
machines.
Node names gpu-pt1-02.math.private
gpu-pt1-03.math.private#Nodes 2 CPU model (2 per node) Intel(R) Xeon(R) CPU E5-2650 v4 @ 2.20GHz #Cores/Node 24 Threads per core 2 System Memory/Node 256 GB GPU Type NVIDIA GeForce GTX 1080ti # GPU devices/Node 8 GPU memory per device 12 GB