Skip to content

Commit

Permalink
only support cardinal cluster (#13)
Browse files Browse the repository at this point in the history
  • Loading branch information
johrstrom authored Oct 31, 2024
1 parent 56a5385 commit c00bc97
Show file tree
Hide file tree
Showing 2 changed files with 25 additions and 34 deletions.
50 changes: 22 additions & 28 deletions form.yml
Original file line number Diff line number Diff line change
@@ -1,9 +1,7 @@
---
cluster:
- "owens"
- "cardinal"
form:
- auto_modules_schrodinger
- schrodinger_version
- auto_accounts
- bc_num_hours
- bc_num_slots
Expand All @@ -13,6 +11,11 @@ form:
- bc_vnc_resolution
- bc_email_on_started
attributes:
schrodinger_version:
widget: 'select'
options:
- ["2023.2", "schrodinger/2023.2"]
- ["2024.2", "schrodinger/2024.2"]
num_cores:
widget: "number_field"
label: "Number of cores"
Expand All @@ -21,48 +24,39 @@ attributes:
Number of cores on node type (4 GB per core unless requesting whole
node). Leave blank if requesting full node.
min: 0
max: 28
max: 96
step: 1
id: 'num_cores'
bc_num_slots: "1"
bc_vnc_resolution:
required: true
node_type:
widget: select
label: "Node type"
help: |
- **any** - (*1-28 cores*) Use any available Owens node. This reduces the
- **any** - (*96 cores*) Use any available Cardinal node. This reduces the
wait time as there are no node requirements.
- **hugemem** - (*4-48 cores*) Use an Owens node that has 1.5TB of
available RAM as well as 48 cores. There are 16 of these nodes on
Owens.
- **vis** - (*1-28 cores*) Use an Owens node that has an [NVIDIA Tesla P100
GPU](http://www.nvidia.com/object/tesla-p100.html) with an X server
- **vis** - (*96 cores*) Use an Cardinal node that has an [NVIDIA H100
GPU](https://www.nvidia.com/en-us/data-center/h100/) with an X server
running in the background. This utilizes the GPU for hardware
accelerated 3D visualization. There are 160 of these nodes on Owens.
accelerated 3D visualization. There are 32 of these nodes on Owens.
- **hugemem** - (*96 cores*) Use an Cardinal node that has 2 TB of
available RAM as well as 96 cores. There are 16 of these nodes on
Owens. Requesting these nodes always reserve entire nodes.
options:
- [
"any", "any",
'any', 'any',
data-max-num-cores: 96,
data-min-num-cores: 1,
data-max-num-cores: 28,
]
- [
"hugemem", "hugemem",
data-min-num-cores: 4,
data-max-num-cores: 48,
data-option-for-cluster-cardinal: false,
'vis', 'vis',
data-max-num-cores: 96,
data-min-num-cores: 1,
]
- [
"vis", "vis",
data-min-num-cores: 1,
data-max-num-cores: 28,
'hugemem', 'hugemem',
data-max-num-cores: 96,
data-min-num-cores: 43,
]
version:
widget: select
label: "Schrodinger version"
help: "This defines the version of Schrodinger you want to load."
options:
- [ "2020.1", "schrodinger/2020.1" ]
licenses:
help: |
If you want to use use macromodel, glide, ligprep, qikprep, or epik feature in Schrodinger,
Expand Down
9 changes: 3 additions & 6 deletions template/script.sh.erb
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
#!/usr/bin/env bash

<%- gpu = context.node_type.include?("vis") -%>
<%- gpu = context.node_type. == "vis" -%>

# Clean the environment
module purge
Expand Down Expand Up @@ -39,14 +39,11 @@ fi
#

# Load the required environment
module load <%= context.auto_modules_schrodinger %>
module load <%= context.schrodinger_version %>

# Launch Schrodinger
<%- if gpu -%>
<%- if context.cluster == 'owens' -%>
module load intel/16.0.3
<%- end -%>
module load virtualgl
module load intel/2021.10.0 virtualgl
module list
set -x
vglrun maestro -NOSGL
Expand Down

0 comments on commit c00bc97

Please sign in to comment.