Here you will find helpful information on how to use the Linux application server, Tesla
Tesla is a Linux-based statistical software server available to researchers and students within the School of Arts & Sciences. While the server provides traditional, x86-based software and programming functionality, it also contains its namesake NVidia Tesla GPU card, for use in GPU-computing applications.
What are Tesla's hardware specifications?
- 2x Intel Xeon E5-2670 @ 2.60GHz (16 cores total)
- 196GB System RAM
- GPU: Nvidia Tesla M2075
- 448 GPU Cores @ 1.15Ghz
- 6GB GDDR5 Memory
- 1 TFlops / 0.52 TFlops Single/Double precision peak performance
- Intel Xeon Phi 7120P
- 61 Cores @ 1.33Ghz
- 16GB Memory
- Supports x86 code
How do I sign up for an account?
You can request an account here. Remember to check the box next to "Linux Account."
What sort of applications are available on Tesla?
- Text-only applications (R, Stata, Stata-MP, Stata-SE)
- Graphical applications (XStata, XStata-MP, XStata-SE, Matlab, SAS)
- Compilers for programmers (GCC, Intel Compilers C/C++ and Fortran, Nvidia CUDA compiler [for GPU coding], PGI Fortran Compiler with CUDA support)
How do I launch applications from the command line in Windows?
You will need to download and install Secure CRT and also Xming, if you wish to use graphical versions of applications (where available). After you have downloaded and installed the necessary software please see instructions here.
How do I launch applications from the command line in Mac OS?
Open Terminal or X11, located in "Utilities." For text sessions, use the command "ssh " (where user is the user name that you requested via the online form). For graphical sessions, use the command "ssh -X ".
X11 is not included with Mountain Lion (see http://support.apple.com/kb/HT5293).
How do I store and access my data?
How do I run tasks in the background?
We recommend using the program Screen, which allows you to start multiple tasks in different console windows that will continue to run after you disconnect