*Agate's 2 login nodes are: agate[2,5]

Job Limits

Each LC platform is a shared resource. Users are expected to adhere to the following usage policies to ensure that the resources can be effectively and productively used by everyone. You can view the policies on a system itself by running:

news job.lim.MACHINENAME

Web version of Agate Job Limits

There are 5 generally available machines on the SCF. Please consider your needs to decide which machine is best for your flavor of work.

  •    Agate—a 48 node machine for up to 36 core jobs (no interconnect)
  •    Mica—a small machine for debugging (8hr job limit)
  •    Jadeita—a machine for small jobs
  •    Magma—a machine for small to medium jobs
  •    Jade—a machine for large jobs (4 jobs/user limit)

Agate is now in General Availability mode and all jobs are scheduled per cpu in two pools.

  •    pdebug— 2 nodes (72 cores), interactive use only.
  •    pbatch—41 nodes (1476 cores), batch use only.
Pools               Max nodes/job       Max runtime
---------------------------------------------------
pdebug                    1(*)           .5 hour
pbatch                    1             200 hours
---------------------------------------------------

NOTE There is a limit of 300 cores per user per bank, aggregated across all jobs on Agate.  If you exceed that limit, squeue will show a status of "AssociationResourceLimit" for your queued jobs.

(*) pdebug is for debugging, visualization, and other interactive work:

  •      Jobs are cpu scheduled
  •      Limit your work to 1 nodes on a per user basis.
  •      Do not use pdebug for production work or chain jobs
  •      Using more than these limits can result in job removal without notice.

Interactive access to a batch node is allowed only while you have a batch job running on that node, and only for the purpose of monitoring your job. When logging into a batch node, be mindful of the impact your work has on the other jobs running on the node.

Multiple users can run on the same node.  If the number of cores is not specified, the job is allocated to one core.

Add this line to your batch script to use 36 cores, and get a node with no other users:

SBATCH SCRIPT: #SBATCH -n 36 or  #SBATCH --exclusive

MSUB SCRIPT: #MSUB -l ttc=36

ALWAYS use the msub -l ttc or sbatch -n option when your job will use more than one task.

Hardware

  •    Intel Xeon E5-2695 2.1GHz cores
  •    Intel Omni-Path Interconnect
  •    36 cores/node
  •    128GB memory/node

Please call or send email to the LC Hotline if you have questions. LC Hotline | phone: 925-422-4531 |   email: lc-hotline@llnl.gov

 

Zone
SCF
Vendor
Penguin
User-Available Nodes
Login Nodes*
2
Batch Nodes
41
Debug Nodes
2
Total Nodes
48
CPUs
CPU Architecture
Intel Xeon E5-2695 v4
Cores/Node
36
Total Cores
1,728
Memory Total (GB)
6,144
CPU Memory/Node (GB)
128
Peak Performance
Peak TFLOPS (CPUs)
58.1
Clock Speed (GHz)
2.1
Peak single CPU memory bandwidth (GB/s)
77
OS
TOSS 4
Interconnect
N/A
Parallel job type
multiple jobs per node
Run Command
srun
Recommended location for parallel file space
Program
ASC
Class
CTS-1
Password Authentication
OTP
Year Commissioned
2016
Compilers

See Compilers page

Documentation