Home research People General Info Seminars Resources Intranet
| Resources | | Computational Resources | New users | Operation guidelines | Downloads | Intranet
Resources | Parasol Laboratory

Parasol Resources Overview

Computational Resources

Computing in the Parasol laboratory is primarily based on the Linux operating system, specifically the community-oriented CentOS, and Open Source software, on both workstation and server platforms. The network services provide version control, project tracking, identity and configuration management, web publishing and database back ends (among other functions).

We also support undergraduate and graduate instruction for our faculty members.


Supercomputing

Cray XE6m Supercomputer at Parasol CRAY XE6m Supercomputer
• 24 compute nodes, with the following characteristics:
  • 12 regular nodes each with two 16 core 64-bit AMD© 'Interlagos' 2.1 GHz CPUs, for a total of 384 cores.
  • 12 accelerated nodes, each with one 16 core 64-bit AMD© 'Interlagos' 2.1 GHz CPU, Kepler© K20X GPU, for a total of 192 cores.
  • Note: Grand total is 576 cores.
  • Additionally, each Kepler K20X has a total of 2496 GPU cores, approaching a peak double precision performance of 1.17 TFLOPS, or 3.52 TFLOPS in single precision. The total GPU core count is 29,952.
• 8 service nodes (boot/login/database/filesys) with 6-core 64-bit AMD 'Istanbul' CPUs.
• CRAY proprietary Gemini 3D torus interconnect fabric.
• CLE (Cray Linux Environment) software stack with GNU, Portland Group©, and CRAY compiler/MPI suites.
• LUSTRE parallel file system with 36TB usable space.
• See the Official Product Page for this system at www.cray.com.

GPU/CUDA High Performance Computing Server

IBM POWER S822LC HPC server at Parasol "Cholesky" Server
• IBM POWER S822LC HPC with 4 NVIDIA Tesla P100 GPU modules.
• Characteristics:
  • Dual 80-core IBM© PPC64LE© (2 threads per core) processors.
  • 4 NVIDIA© Tesla© P100, 3584 cores each, providing a total of 14336 GPU cores.
  • Benchmarked to 5.73 TFLOPS (double precision), 10.6 TFLOPS (single precision).
  • 1 TB RAM
  • 1 TB disk (SSD) scratch space
  • Operating System: Red Hat Enterprise Linux 7
  • Software: CUDA, MATLAB, AMPL, and other mathematical modeling packages.

GPU/CUDA High Performance Computing Server

The Supermicro SuperServer 4027GR-TRT HPC GPU server at Parasol "Backslash" Server
• Supermicro "Mercury" AH-GPU424 HPC server with 8 NVIDIA Tesla K40m GPU modules.
• Characteristics:
  • Dual 12-core Intel© Xeon© CPU E5-2695 v2 @ 2.40GHz (2 threads per core) processors.
  • 8 NVIDIA© Tesla© K40m, 2880 cores each, providing a total of 23040 GPU cores.
  • Benchmarked to 1.78 TFLOPS (double precision), 3.0 TFLOPS (single precision).
  • 768 GB RAM
  • 8 TB disk scratch space
  • Operating System: CentOS 7
  • Software: CUDA, MATLAB, AMPL, and other mathematical modeling packages.

Personal Workstations



Parasol Home | Research | People | General info | Seminars | Resources  

Parasol Laboratory, 425 Harvey R. Bright Bldg, 3112 TAMU, College Station, TX 77843-3112 
parasol-admin@cse.tamu.edu      Phone 979.458.0722     Fax 979.458.0718 

Department of Computer Science and Engineering | Dwight Look College of Engineering Dwight Look College of Engineering | Texas A&M University
    
Privacy statement: Computer Science and Engineering Engineering TAMU
Web Accessibility Policy and Law - Web Accessibility and Usability Standards  -   Contact Webmaster