113
edits
Difference between revisions of "Computer Info"
Jump to navigation
Jump to search
→Computer System Information for the Math Department
|
The Math Department has a separate computer account system for logging in to Linux workstations and the computation machines. This allows faster operations in many cases. It also protects users' netid accounts because Math Department accounts have no ability to access resources under NetID accounts. To set up a Math account, a Math instructor or system administrator must send an 'account invite' to a user's NetID. Once the user has been invited, they can use their NetID account and two-factor authentication to access the Math system and set up their Math account. To change the password on a Math department account, or to invite a user to set up a Math account, go to [https://accounts.math.cornell.edu Math Accounts Page]
Math department account files are stored on a central file server, and accessed over the network by the workstations. These files are available to the computation machines over a high-speed network in the server room.
'''Linux
The Math Department has Linux workstations which can be used for general computer use, as well as for access to Math computation tools and licensed commercial Math software, and the computation resources. They also have a full array of open-source software as well as numerous LaTeX tools for creating and working with Math documents.
'''Linux Computation Machines'''
Math has several machines dedicated to computation. They can be accessed by ssh or by using remote desktop software. The environment is standardized between workstations and computation machines so that the user will have an environment that is familiar to them. The computation machines have different strengths, from GPU workstations to machines with very large memory and core counts. You can view the [[Machine List]] here.
General-purpose GPU computing.▼
The computation machines are housed in a special server room which has backup power, high-capacity cooling, and a high-speed network.
Linux HPC cluster.▼
▲'''General-purpose GPU computing.'''
We have several GPU workstations for shared remote use. These machines have Nvidia RTX 2080 or RTX 3080 GPU units in them. They support CUDA and other libraries, and software including Tensorflow and other tools. All of our commercial Math packages have support for GPU calculations. Matlab, Mathematica, Maple, and Magma are all supported on our GPU systems.
▲'''Linux HPC cluster.'''
'''Macintosh computers.'''
| |||