Systems and Software

This page lists information about the system and available software, languages, compilers, modules, etc. This is only a partial list, so if there is anything you are interested in that isn't listed here, please contact us.

MGHPCC TX-E1 Specifications

Summary
Number of Nodes 704
Total CPU Cores 32000
Total GPUs 448
Distributed Storage 873 TB
Individual Nodes
Processor Nodes CPU Cores Node RAM RAM per core GPU Type GPUs
(per Node)
GPU RAM Local Disk
Intel Xeon
Gold 6248
224 2 x 20 384 GB 9 GB Nvidia Volta V100 2 32 GB 3.8 TB
Intel Xeon
Platinum 8260
480 2 x 24 192 GB 4 GB NA NA NA 4.4 TB
Resource Allocations
Processor Partition Starting Standard
Intel Xeon
Gold 6248
xeon-g6-volta 1 Nodes
40 Cores
2 GPUs
6 Nodes
240 Cores
12 GPUs
Intel Xeon
Platinum 8260
xeon-p8 2 Nodes
96 Cores
16 Nodes
768 Cores

Available Languages

  • Julia
  • Python (Anaconda available through modules)
  • Matlab(R)/Octave
  • R
  • C/C++
  • Fortran
  • Java
  • Perl 5
  • Ruby

Software

Machine Learning Tools

  • Tensorflow
  • Pytorch

Big Data Software Stack

  • Hadoop
  • Zookeeper
  • Accumulo

Middleware Software Stack

  • ARPACK
  • ATLAS
  • Boost
  • BLAS
  • FFTW
  • LAPAC
  • OpenMPI
  • OpenBLAS

Lincoln Laboratory Developed Software

Compilers

  • gcc
  • g++
  • gfortran
  • icc

Modules

To see the most up-to-date list of currently available modules, run the command module avail. For more information about modules, see the module section on the Software and Package Management page