compute_ru

User Tools

Site Tools

Trace:
language_lab:cluster

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision		Previous revision
language_lab:cluster [2023/03/24 11:44] – bjornslanguage_lab:cluster [2024/10/14 14:24] (current) – external edit 127.0.0.1
Line 1: Line 1:
-======Smallvoice - the Language and Voice lab computing cluster======+======SoT cluster - the School of Technology computing cluster======
 \\ \\
  **Smallvoice**, uses [[https://slurm.schedmd.com|Slurm]] workload manager to create a computing cluster\\  **Smallvoice**, uses [[https://slurm.schedmd.com|Slurm]] workload manager to create a computing cluster\\
-The Cluster has 6 nodes :\\ 
-^Node name ^ Role(s) ^ 
-|**atlas**|management node| 
-|**freedom**|**login** node, worker| 
-|hercules|worker node| 
-|samson|worker node| 
-|goliath|worker node| 
-|obelix|worker node| 
  
-When logged on to the cluster user is always on the login node, **freedom** and does all his work there.\\ /home (& work) are hosted on a NFS server, so every nodes have the **same** "physical" disks\\+When logged on to the cluster user is always on the login node, called **freedom** and should do all his work there.\\ Home folder for all users are hosted on a NFS server, so every nodes have the **same** "physical" disks\\ 
 +All user-jobs should run using slurm sbatch job.sh, please do not run job locally on the login node\\
 \\ \\
 === The computing (slurm) environment === === The computing (slurm) environment ===
-There are 3 partitions / queue's available\\ +There are 2 partitions (queue's)\\ 
-^Name^Cores^Memory(gb)^Nodes^GPU^Timelimit^Usage^ +^Name^Nodes^GPU^Timelimit^ 
-|allWork|16+18+12+12|64+40+48+48|4|Nvidia A100 GPU|7 days|staff only| +|basic|3|Nvidia A100 GPU|31 hours| 
-|doTrain|16+18+12+18|64+40+48+40|4|Nvidia A100 GPU|no limit|staff only| +|lvlWork|3|Nvidia A100 GPU|no limit|
-|beQuick|18+12|40+48|2|Nvidia A100 GPU|36 hours|for students| +
-|bigVoice|12|32|1|Nvidia A100 GPU|no limit|staff|+
  
-The **default** queue for staff is doTrain (and beQuick for student) so it's not necessery to choose a queue, but it's possible to specify a different one.\\+ 
 +The **default** queue for both student and staff is basic, so it's not necessery to choose a queue in your script file, but it's possible to specify a different one.\\
  
 ===Installed software and drivers=== ===Installed software and drivers===
  * **NVIDIA** A100 GPU drivers\\  * **NVIDIA** A100 GPU drivers\\
  * **CuDA toolkit** [version 11.7]\\  * **CuDA toolkit** [version 11.7]\\
- * Intel oneAPI Math Kernel Library\\ + * Python 3.9.7\\
- * Python 3.9.2\\+
  * pip 20.3.4\\  * pip 20.3.4\\
  * ffmpeg + sox  * ffmpeg + sox
  
-If additional software is needed or different version, you can ask sysadmin (compute@ru.is) for assistance\\+If additional software is needed or different version, you can ask sysadmin (help@ru.is) for assistance\\
language_lab/cluster.1679658295.txt.gz · Last modified: 2024/10/14 14:24 (external edit)