SOSCIP GPU
| P8 | |
|---|---|
| Installed | September 2017 |
| Operating System | Ubuntu 16.04 le |
| Number of Nodes | 14x Power 8 with 4x NVIDIA P100 |
| Interconnect | Infiniband EDR |
| Ram/Node | 512 GB |
| Cores/Node | 2 x 10core (20 physical, 160 SMT) |
| Login/Devel Node | sgc01 |
| Vendor Compilers | xlc/xlf, nvcc |
SOSCIP
The SOSCIP GPU Cluster is a Southern Ontario Smart Computing Innovation Platform (SOSCIP) resource located at theUniversity of Toronto's SciNet HPC facility. The SOSCIP multi-university/industry consortium is funded by the Ontario Government and the Federal Economic Development Agency for Southern Ontario [1].
Specifications
The SOSCIP GPU Cluster consists of of 14 IBM Power 822LC Servers each with 2x10core 3.25GHz Power8 CPUs and 512GB Ram. Similar to Power 7, the Power 8 utilizes Simultaneous MultiThreading (SMT), but extends the design to 8 threads per core allowing the 20 physical cores to support up to 160 threads. Each node has 4x NVIDIA Tesla P100 GPUs each with 16GB of RAM with CUDA Capability 6.0 (Pascal) connected using NVlink.
Compile/Devel/Test
Access is provided through the BGQ login node, bgqdev.scinet.utoronto.ca node using ssh, and from there you can proceed to the development nodesgc01.
Filesystem
The filesystem is shared with the BGQ system. See here for details.
Software
GNU Compilers
To load the newer advance toolchain version use:
module load gcc/6.2.1
IBM Compilers
To load the native IBM xlc/xlc++ compilers
For p8t0[3-4]
module load xlc/13.1.5 module load xlf/13.1.5
Driver Version
The current NVIDIA driver version is 361.93
CUDA
The current installed CUDA Tookit is 8.0
module load cuda/8.0
The CUDA driver is installed locally, however the CUDA Toolkit is installed in:
/usr/local/cuda-8.0
OpenMPI
Currently OpenMPI has been setup on the four nodes connected over EDR Infiniband.
$ module load openmpi/1.10.3-gcc-6.2.1 $ module load openmpi/1.10.3-XL-13_15.1.5
OpenAI
The IBM OpenAI framework is installed.
Job Submission
The SOSCIP GPU cluster uses SLURM as a job scheduler and jobs are scheduled by node, ie 20 cores and 4 GPUs.