CQLS Oregon State University CQLS Account Request About Policies ACTF Account Request CQLS Support Request Oregon State University

Donations to the CQLS HPC at OSU

Main Page Documentation Donations Software Mail lists CQLS Services Cloud Status CQLS Policies Announcements

New AMD Server with Radeon Instinct GPU

  • The CGRB/CQLS working with Tom Raisor and Travis Campbell at
    Advanced Micro Devices (AMD) to bring online a new computational
    server that will be used to support the Monarch Initiative ( https://monarchinitiative.org/)
    and the Translational and Integrative Sciences Lab ( https://tislab.org/) on COVID-19 research.


  • Hardware: 1x Supermicro AS-4124GS-TNR 4U Rackmount Server (each server has 2 socket x 64 core x 2 thread)
  • Processor: AMD EPYC 7662
  • Memory: 1024GB DDR4 2933 MT/s Buffered RAM
  • Dual Port 40G Ethernet (each server)
  • 4x 960GB SSD (each server)
  • 8x AMD Radeon Instinct MI-50 32GB
  • Date: June 2020
  • Server(s):
  • Lab: CGRB/CQLS and TIS Lab
  • Contact: chris at cgrb.oregonstate.edu
  • Manufacturer: AMD

New Cluster Files Space

  • The CGRB/CQLS working with Mark Lawrence and Björn Kolbeck at
    Quobyte and Joe Lipman at Advanced HPC, Inc. to bring online the new computational
    infrastructure cluster file space used for researchers to work with large amounts of data
    at high speed and low cost overall cost. This space is used for our high speed long term
    storage and capturing all research data from 26 departments across Oregon State
    University. We are currently generating 4-8TB of data per day coming from a gamut
    of sources and needed a file space that could keep up while reducing costs
    around labor to manage. After testing many configurations and different
    groups Quobyte was the only group that could meet all requirments and
    reduce labor at the same time. Home directory space is handled by other
    file services to separate our functions and provide the most robust HPC solution.
  • Hardware: 4x Mercury RM436 4U Rackmount Server (Supermicro servers each server has 2 socket x 4 core x 2 thread)
  • Memory: 96GB
  • Dual Port 12G/s SAS HBA (each server)
  • Dual Port 40G Ethernet (each server)
  • 2x 250GB SSD (each server)
  • 36x 14TB SATA (each server)
  • Date: June 2019
  • Server(s): Deployment of Quobyte with starting 1+PB to replace our current 5PB of space on ZFS and ability to grow as needed for reasearchers at Oregon State.
  • Lab: CGRB/CQLS Oregon State
  • Contact: chris at cgrb.oregonstate.edu
  • Manufacturer: Quobyte

  • The CGRB/CQLS working with David Selle at
    Nyriad and Joe Lipman at Advanced HPC, Inc. to bring online the new computational
    infrastructure home directory file space used for researchers to work within our
    HPC environment. This space is used for our day to day operations for all researchers
    home directory data from 26 departments across Oregon State University. After testing
    many configurations and different groups Nyriad was able to meet all requirments
    needed around really fast home directory space with incredible rebuild times. This
    configuration was able to reduce impact to users and all interaction across the
    infrastructure when hardware failed ensuring users were able to work without
    slowing down. Large research file space is handled by other file services to
    separate our functions and provide the most robust HPC solution.
  • Hardware: 1x NSULATE 4U Rackmount Server
  • Memory: 96GB
  • Internal 12G/s SAS HBA
  • Dual Port 40G Ethernet
  • 1x 250GB SSD (each server)
  • 12x 14TB SATA (each server)
  • Date: June 2019
  • Server(s): Deployment of NSULATE with starting 50+TB to replace our current home directory space with the ability to grow as needed for reasearchers at Oregon State.
  • Lab: CGRB/CQLS Oregon State
  • Contact: chris at cgrb.oregonstate.edu
  • Manufacturer: Nyriad

  • The CGRB/CQLS working with Stan Gowen at
    IBM to provide the computational infrastructure a new POWER9
    Newell processing server.
     
  • Hardware: 2x POWER S922LC Newell servers with 2x Power9 processors each (each server has 2 socket x 8 core x 10 thread = 160 thread)
  • Memory: 1024GB
  • 4x V100 GPU
  • Dual Port 40G Ethernet
  • 2x 3.2TB NVMe CAPI Flash Cards
  • Date: May 2018
  • Server(s): Continuous development environment for users to
    test and build code on Power8 systems. Access request form locate at "http://osuosl.org/services/powerdev/request_gpu"
  • Lab: CGRB/CQLS Oregon State
  • Contact: chris at cgrb.oregonstate.edu
  • Manufacturer: IBM
  • Software Compiled on Power

  • The CGRB/CQLS working with Stan Gowen at
    IBM to provide the computational infrastructure a new POWER9
    Newell processing server.
     
  • Hardware: 2x POWER S922LC Newell servers with 2x Power9 processors each (each server has 2 socket x 8 core x 10 thread = 160 thread)
  • Memory: 1024GB
  • 4x V100 GPU
  • Dual Port 40G Ethernet
  • 2x 3.2TB NVMe CAPI Flash Cards
  • Date: May 2018
  • Server(s): Continuous development environment for users to
    test and build code on Power8 systems. Access request form locate at "http://osuosl.org/services/powerdev/request_gpu"
  • Lab: CGRB/CQLS Oregon State
  • Contact: chris at cgrb.oregonstate.edu
  • Manufacturer: IBM
  • Software Compiled on Power

 

New High Thread Count System for HPC/Biocomputing

 

New GPU for Deeplearning

  • The CGRB/CQLS working with John Saposhnik at NVIDIA to provide the taching computational infrastructure two new NVIDIA GPU processor.
     
  • Hardware: 2xK80
  • Memory: 12G
  • Date: Feburary 2017
  • Server(s): ACTF GPU Teaching Server - On the Advanced Cyberinfrastructure Teaching Facility (ACTF), learners and instructors explore computation at the big-data end of the spectrum, in topics ranging from Bioinformatics to Data Science.
  • Lab: ACTF - CGRB/CQLS Oregon State
  • Contact: chris at cgrb.oregonstate.edu
  • Manufacturer: NVIDIA

 

New Development Environment

  • The CGRB/CQLS working with Charlie Foretich at
    IBM to provide the computational infrastructure a new POWER
    Minsky processing server.
     
  • Hardware: 3x POWER S822LC Minsky servers with 2x Power8 processors each (each server has 2 socket x 8 core x 10 thread = 160 thread)
  • Memory: 1024GB
  • 2x P100 GPU
  • Dual Port 40G Ethernet
  • 2x 3.2TB NVMe CAPI Flash Cards
  • Date: May 2017
  • Server(s): Continuous development environment for users to
    test and build code on Power8 systems. Access request form locate at "http://osuosl.org/services/powerdev/request_gpu"
  • Lab: CGRB/CQLS Oregon State
  • Contact: chris at cgrb.oregonstate.edu
  • Manufacturer: IBM
  • Software Compiled on Power

 

New GPU for Deeplearning

  • The CGRB/CQLS working with Bob Crovella at NVIDIA to provide the computational infrastructure a new NVIDIA GPU processor.
     
  • Hardware: K40
  • Memory: 12G
  • Date: Feburary 2016
  • Server(s): Processing of data using the Kaggle National Data Science Bowl Deep Sea Learning alogrythm to

    Predict ocean health, one plankton at a time.

  • Lab: Cowen Lab Oregon State
  • Contact: chris at cgrb.oregonstate.edu
  • Manufacturer: NVIDIA

 

New Processing Server

  • The CGRB/CQLS working with Charlie Foretich at
    IBM to provide the computational infrastructure a new POWER S822LC processing server.
     
  • Hardware: POWER S822LC
  • Memory: 512GB
  • K80 GPU
  • Dual Port 40G Ethernet
  • Date: December 2015
  • Server(s): Processing of computational biology, high throughput sequencing, analytical chemistry and other computational methods.
  • Lab: CGRB/CQLS Oregon State
  • Contact: chris at cgrb.oregonstate.edu
  • Manufacturer: IBM
  • Software Compiled on Power

 

Upgraded Network

  • The CGRB/CQLS working with David M. Björk from Dell provided a Force10 s6000 network switch with 32 ports of 40Gbps Ethernet. This setup has become the main distrobution switch for the CGRB/CQLS file servers.
  • Hardware: Force10 s6000
  • Date: Fall 2014
  • Server(s): All file servers connect through this equipment
  • Lab: CGRB/CQLS Oregon State
  • Contact: chris at cgrb.oregonstate.edu
  • Manufacturer: Dell

 

Upgraded Network

  • The CGRB/CQLS working with David M. Björk from Dell provided a Force10 s4810 network switch with 48 ports of 10Gbps Ethernet and 4 ports of 40Gbps Ethernet. This setup has become the main distrobution switch for the CGRB/CQLS file servers.
  • Hardware: Force10 s4810
  • Date: Fall 2012
  • Server(s): All file servers connect through this equipment
  • Lab: CGRB/CQLS Oregon State
  • Contact: chris at cgrb.oregonstate.edu
  • Manufacturer: Dell

 

Plant Ontology Consortium (POC) / Gramene: A Resource for Comparative Grass Genomics

 

Assembling the Fungal Tree of Life

  • The Spatafora and CGRB/CQLS working with Kasey Flicker from MCPc and Benjamin Benulis from AMD (Advanced Micro Devices) have setup a 20 Opteron AMD processors for working with the Fungal Tree of Life project.
  • Hardware: 20 AMD Opteron(tm) Processor 246
  • Date: March 2005
  • Node(s): fungi1-10
  • Website: http://aftol.org
  • Lab: Dr. Joseph Spatafora Lab Oregon State
  • Contact: spatafoj at science.oregonstate.edu
  • Manufacturer: MCPc (Kasey Flicker Senior Account Manager)

 

HPC Cluster and Bioinformatics