You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: docs/userguides/scholar/scholar-spark.md
+10-13Lines changed: 10 additions & 13 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -6,12 +6,11 @@ authors:
6
6
date:
7
7
created: 2026-06-11
8
8
updated: 2026-06-11
9
-
search:
10
-
exclude: true
11
9
resource: Scholar
12
10
host: scholar.rcac.purdue.edu
13
11
search:
14
12
boost: 2
13
+
exclude: true
15
14
---
16
15
17
16
# Scholar Spark Nodes
@@ -64,16 +63,6 @@ The Spark nodes are are split into two partitions:
64
63

65
64
66
65
67
-
### ECE Spark Nodes
68
-
69
-
<!-- !!! question
70
-
**TODO** If RCAC + ECE nodes are in the same interactive and batch partitions (per slack), I'm assuming that access to ECE nodes is managed by the reservations (as per the scholar spark expansion word doc)? -->
71
-
72
-
73
-
The 80 DGX Spark nodes within the `scholar-k` subcluster are for exclusive use by ECE researchers and courses at the Purdue Main Campus and Purdue Indianapolis. All eligible ECE users are added to reservations that will grant them access to these nodes, and will be automatically used for submitted jobs.
74
-
75
-
Of these, 10 (`scholar-kxxx`-`scholar-kxxx`) are exclusively for use by Purdue Main Campus, 10 (`scholar-kxxx`-`scholar-kxxx`) are exclusively for use by Purdue Indianapolis, with the remaining nodes existing within a shared pool.
76
-
77
66
78
67
## Accessing Spark Nodes
79
68
@@ -334,10 +323,18 @@ module load modtree/spark
334
323
335
324
If you have built or downloaded your own applications, please rebuild any applications on a Spark node, or check for an `aarch64` distribution of your software.
336
325
326
+
### How can I check the architecture of the node I'm running on?
327
+
328
+
Since Scholar has nodes with `x86_64` architecture and `aarch64` architecture, it is important to pay attention to what the architecture is on the node you plan to build or run software on. You can use the following command to print the architecture of the node you are currently running on.
329
+
330
+
```bash linenums="0"
331
+
uname -m
332
+
```
333
+
337
334
338
335
### How can I separate my `x86_64` applications from `aarch64` applications?
339
336
340
-
If you plan on working on both `aarch64` and `x86_64` nodes, we strongly reccomend separating applications and configurations in separate directories.
337
+
If you plan on working on both `aarch64` and `x86_64` nodes, we strongly recommend separating applications and configurations in separate directories.
341
338
342
339
For example, you may choose to make separate `bin` directories for `x86_64` and `aarch64` applications. In your `~/.bash_profile` file, you can then source the correct applications based on the architecture:
0 commit comments