Ares Computer Cluster
Ares is a high-performance computing cluster shared between the Cao, Ehmann, Liu, Wagner, Chen, Keten, and Packman groups. Many commercial softwares have the option to be run in parallel, which greatly improves simulation speed.
Features:
-
Ares consists of 32 nodes with 24 cores per node (with 64Gb of memory per node) for a total of 768 CPUs
-
Software installed includes:
-
Abaqus
-
Bolt
-
ImageJ/Fiji
-
LAMMPS
-
LS-DYNA
-
Matlab
-
NAMD
-
VMD
-
and various debuggers and compilers
-
We have a user wiki and a real-time monitoring system (Ganglia) for Ares here: http://ares.mech.research.northwestern.edu/
AMPL Process Analysis Workstation
Advanced Manufacturing Processes Laboratory Process Analysis Workstation (AMPL PA) is a computational resource available to all students, staff and visiting scholars in AMPL.
Features:
AMPL PA workstation has 40 CPU cores, 256 GB of RAM, and NVIDIA Quadro K6000 GPU running on Windows 10 Enterprise edition OS.
Software installed includes:
- ABAQUS
- SOLIDWORKS
- COMSOL Multiphysics
- Anaconda
- Visual Studio 2015
- MATLAB
The purpose of AMPL PA is to facilitate development and optimization of various manufacturing processes and machine designs with the help of virtual modeling tools and numerical simulations of the phenomena related to the manufacturing processes.
AMPL Leonidas and Yoda Workstations
AMPL Leonidas and Yoda are two dedicated computational workstations available to all students and researches associated with AMPL. The purpose of AMPL Leonidas and Yoda workstations is to accelerate computationally expensive simulations, and to make essential commercial software available to AMPL researchers.
Features:
Leonidas: 32 cores and 74 GB of RAM running Red Hat Enterprise Linux Server release 6.9
Yoda: Dual Intel Xeon 24C CPU and 128 GB of RAM running on Red Hat Enterprise Linux 7.3
Software installed includes:
- ABAQUS
- LS_DYNA
- MATLAB
- COMSOL
- LS-OPT
- CMAKE
- Various compilers
AMPL Office Workstation
AMPL Office workstation is a computational resource located in B110 office available to all students and researches associated with AMPL. This workstation provides multiple software packages to facilitate design, development, and prediction of manufacturing processes.
Features:
AMPL Office Workstation has Intel Xeon 8C CPU, 64 GB of RAM, and NVIDIS Quadro P5000 GPU running on Windows 10 Pro OS.
Software installed includes:
- ABAQUS
- MATLAB
- Anaconda
- Microsoft Office
- Adobe package
- Visual Studio 2017
Lambda Quad GPU Workstations
AMPL is equipped with three Lambda Quad GPU workstations: Darth Vader, Darth Maul and Leia. All AMPL associated students and scholars are welcome to use it for their graphics-enabled computational works such as CUDA programs and deep learning scripts with Tensorflow or Pytorch.
Features:
Darth Vader workstation has four Nvidia RTX 8000 GPUs (each with 4608 CUDA cores and 48 GB of memory), 256 GB RAM and an Intel Xeon W-2195 CPU with 18 processing cores running on Ubuntu 18.04.
Darth Maul workstation has four Nvidia RTX 8000 GPUs (each with 4608 CUDA cores and 48 GB of memory), 256 GB RAM and an AMD Threadripper 3970X CPU with 32 processing cores running on Ubuntu 20.04.
Leia workstation has four Nvidia RTX A6000 GPUs (each with 4608 CUDA cores and 48 GB of memory), 256 GB RAM and an AMD Threadripper Pro 3975WX CPU with 32 processing cores running on Ubuntu 20.04.
Installed software and packages include:
- TensorFlow
- PyTorch
- Caffe
- Keras
- CUDA
- cuDNN
