ARCS to absrocks/Docs

.github/workflows/static.yml

@@ -3,12 +3,14 @@ on:
   push:
     branches:
       - main
-      - alex_patch 
-      - abhi_patch
+      - ARCS-Lakshya
+      - ARCS-Abhi
+      - julia_patch      
     workflow_dispatch:
       - main
-      - alex_patch
-      - abhi_patch
+      - ARCS-Lakshya
+      - ARCS-Abhi
+      - julia_patch
 permissions:
   contents: write
 jobs:
@@ -26,5 +28,5 @@ jobs:
           path: .cache
           restore-keys: |
             mkdocs-material-
-      - run: pip install mkdocs-material pandas tabulate markdown-exec mkdocstrings mkdocs-table-reader-plugin mkdocs-awesome-pages-plugin mkdocs-autolinks-plugin
+      - run: pip install mkdocs-material pandas tabulate markdown-exec mkdocstrings mkdocs-table-reader-plugin mkdocs-awesome-pages-plugin mkdocs-autolinks-plugin mkdocs-literate-nav
       - run: mkdocs gh-deploy --force

.gitignore

@@ -1,2 +1,3 @@
 # MkDocs documentation
-site*/
+site*/
+.DS_Store

docs/Courses/course-job-submission.md

@@ -0,0 +1,100 @@
+# Submitting Course Jobs
+
+Submit scripts on Wulver must include specification of partition, account, qos, and time limit. Also the first line in the batch file must be `#!/bin/bash -l`. 
+
+Below is the minimal example for a 10-minute test job:
+
+```bash
+#!/bin/bash -l
+#SBATCH --partition=course
+#SBATCH --account=2025-fall-ds-492-kjc59-ls565
+#SBATCH --qos=course
+#SBATCH --time=00:10:00
+```
+
+### Sample Job scripts
+
+Make sure to replace `--account` with your assigned course account.
+
+#### CPU Job Example
+
+```bash
+#!/bin/bash
+#SBATCH --job-name=mpi_test_job
+#SBATCH --output=%x.%j.out
+#SBATCH --error=%x.%j.err
+#SBATCH --partition=course
+#SBATCH --account=2025-fall-ds-492-kjc59-ls565
+#SBATCH --qos=course
+#SBATCH --time=00:10:00
+#SBATCH --ntasks=64
+
+# Run application commands
+srun /apps/testjobs/bin/mpihello
+```
+
+- Runs an MPI job named mpi_test_job.
+- Uses 64 processes across available nodes.
+- Wall time: 10 minutes.
+
+#### GPU Job Example
+
+```bash
+#!/bin/bash
+#SBATCH --job-name=test_gpu_job
+#SBATCH --output=%x.%j.out
+#SBATCH --error=%x.%j.err
+#SBATCH --partition=course_gpu
+#SBATCH --account=2025-fall-ds-492-kjc59-ls565
+#SBATCH --qos=course
+#SBATCH --time=00:20:00
+#SBATCH --ntasks=2
+#SBATCH --gres=gpu:a100_10g:1
+
+# Load application environment
+module load CUDA
+
+# Run application commands
+nvidia-smi
+```
+
+- Runs a GPU job named test_gpu_job.
+- Allocates 2 CPUs and single A100 10G MIG GPU.
+- Wall time: 20 minutes.
+
+### Limitation of GPU Jobs
+
+- You cannot request multiple MIG instances in one job. <br>
+For example: `--gres=gpu:a100_10g:2` <br>
+This will either cause an error or misinterpretation as a single GPU.
+
+- Each job should request one GPU per job. <br>
+For multiple tasks, use job arrays instead of multiple MIGs.
+
+
+### Interactive jobs
+
+You can also start an interactive session instead of a batch job.
+
+```bash
+interactive -a ACCOUNT -q QOS -p PARTITION -j JOB_TYPE
+```
+
+Parameters:
+
+- `-a ACCOUNT`  → Your assigned course account
+- `-q course`   → QoS for course jobs
+- `-p course`   → Partition (use course or course_gpu)
+- `-j JOB_TYPE` → Type of job (e.g., cpu or gpu)
+
+Example
+
+```bash
+interactive -a 2025-fall-ds-492-kjc59-ls565 -q course -p course
+```
+
+This command launches a temporary compute session for hands-on work or testing.
+
+
+!!! info "Learn More About Job Submission"
+    For more detailed examples and advanced options, visit the [**Running Jobs** page](../Running_jobs/index.md).

docs/Courses/course-resource-config.md

@@ -0,0 +1,116 @@
+# Course Resource Configuration
+
+## Overview
+
+Before submitting a job on Wulver, it’s important to understand how resources such as **cores**, **memory**, and **GPUs** are allocated for course users.  
+Each job’s runtime and Service Unit (SU) charge depend on how many nodes and resources are requested.
+
+![Course Partition Overview](../assets/images/course-config-schematic.png){ width=80% }
+
+## Partition (Use `--partition`)
+
+Course-related jobs must be submitted to one of the **course partitions** listed below:
+
+| Partition | Nodes | Cores per Node | GPU | Memory per Node | SU Charge per Hour |
+|------------|:------:|:------------:|:----:|:---------:|:-----------|
+| <code>--&#8203;partition&#8203;=&#8203;course</code> | 64 | 32–48 | NA | 375 GB | `MAX(CPUs, Memory/4G) SU` |
+| <code>--&#8203;partition&#8203;=&#8203;course_gpu</code> | 4 | 32–48 | A100 10G MIG | 375 GB | `2 + MAX(CPUs, Memory/4G) SU` |
+
+Courses that require GPU computing are assigned **NVIDIA A100 10G MIG** units for GPU-enabled coursework.
+
+## Service Units
+
+Each user is allocated **2500 Service Units (SUs)** in their account for the duration of their course.  
+SUs are consumed based on how many cores, memory, and GPUs your job requests and how long it runs.
+
+#### SU Calculation Examples
+
+**Example 1 – CPU-only job** <br>
+20 cores for 8 hours (no `--mem` specified, so default 4G per core will be allocated) 
+```
+SU = MAX(CPUs, Memory/4G) x Hours 
+SU = (20, 4/4) × 8 = 160  
+```
+
+**Example 2 – GPU job**  <br>
+20 cores + 1 GPU for 8 hours with `--mem=128G` <br>
+```
+SU = [2 + MAX(CPUs, Memory/4G)] x Hours 
+SU = [2 + MAX(20, 128/4)] × 8 
+SU = (2 + 32) × 8 = 34 × 8 = 272
+```
+
+*To make the most of your SUs, avoid overestimating cores or memory. Over-allocation wastes SUs and delays scheduling.*
+
+Please visit [here](../Running_jobs/service-units.md) for more info on Service Units
+
+## Account
+
+Every course is assigned a dedicated **account** for SU billing.  
+You must specify this account in all SLURM job scripts using the `--account` flag.
+
+Account Format
+```bash
+<semester>-<course>-<instructor_ucid>-<student_ucid>
+```
+
+For example:
+```bash
+#SBATCH --account=2025-fall-ds-492-kjc59-ls565
+
+# 2025-fall → Semester
+# ds-492 → Course code
+# kjc59 → Instructor’s UCID
+# ls565 → Student’s UCID
+```
+
+!!! note
+    Each user’s specific account information will be emailed to them when HPC resources for that course is allocated. You can also view all accounts assigned to you using the `quota_info` command.
+
+#### Important Notes:
+- Your UCID and password are used to log in to the cluster.
+- Your account name is used for SU tracking and billing.
+- If you are enrolled in multiple courses or research projects, ensure that you submit jobs using the correct account to avoid suspension or access issues.
+- Use the `quota_info` command to view all accounts assigned to you.
+
+
+## Priority (Use `--qos`)
+
+All course-related jobs must include the `--qos=course` flag.  
+This ensures fair scheduling and appropriate priority for classwork jobs.
+
+| Qos | Purpose | Wall time limit (hours) | Valid Users | 
+|------------|:------|:----|:---------|
+| <code>--&#8203;qos&#8203;=&#8203;course</code> | For all course-related jobs | 72 | Course users only |
+
+Example:
+```bash
+#SBATCH --qos=course
+```
+
+## Course Directory
+
+Each student is assigned a dedicated **course directory**, and **all work must be done inside it**, including job scripts and outputs.
+
+Directory Format
+```bash
+/course/<year>/<semester>/<course>/<instructor_ucid>/<student_ucid>
+```
+
+Example:
+```bash
+/course/2025/fall/ds/492/kjc59/ls565
+```
+
+To switch to your course directory after logging in:
+```bash
+cd /course/2025/fall/ds/492/kjc59/ls565
+```
+
+Or add it to your `.bash_profile` for automatic navigation:
+```bash
+echo 'cd /course/2025/fall/ds/492/kjc59/ls565' >> ~/.bash_profile
+```
+
+!!! note
+    Your specific course directory path will also be emailed to you along with account information when resources are allocated.

docs/Courses/index.md

@@ -0,0 +1,58 @@
+# HPC Resources for Teaching & Coursework
+
+## Introduction
+Instructors can utilize High-Performance Computing (HPC) resources for academic courses. Whether faculty are planning a course that involves computationally intensive tasks or introducing students to parallel computing concepts, the HPC environment can offer valuable resources.
+
+## Course Request
+To request HPC resources for course, you need to provide the following information at [HPC Course Request Form](https://nexus.njit.edu/highlander_nexus?id=sc_cat_item&sys_id=cd6eaea13b97e210e914eb0864e45a7f). 
+
+### Name of Course:
+[Please provide the name of the course and a short description]
+
+### Is this one section of a larger course?
+[Yes or No]
+
+### Estimated Number of Students:
+[Enter the estimated number of students for the course]
+
+### Activities on HPC:
+[Describe the specific activities that will involve the use of HPC resources. For example, simulations, data analysis, modeling, etc.]
+
+### Software Needed:
+[Specify the software required for the course. Check the list of software installed on Wulver in [Software](../Software/index.md#software-list). Include any specific versions or software not already available on the HPC cluster] 
+
+!!! warning "Notice Period"
+
+    A minimum of 30 days' notice is required for requesting specific software installations or substantial resource allocations.
+
+## When to Use HPC in a Course
+HPC resources are good for the courses if they satisfy the following requirements.
+
+* **Simulations and Modeling**:
+Perform complex simulations and modeling exercises that require significant computational power.
+* **Data Analysis**:
+Conduct large-scale data analysis projects, exploring real-world datasets with efficiency.
+* **Parallel Computing**:
+Teach parallel computing concepts and applications by leveraging the cluster's parallel processing capabilities.
+* **Optimization Problems**:
+Solve optimization problems that benefit from parallel processing and distributed computing.
+* **Scientific Research Projects**:
+Enable students to work on scientific research projects that demand high-performance computing resources.
+
+## HPC Introduction for Courses
+
+The **HPC Facilitator** is available to provide an introduction to High-Performance Computing. This introduction can be conducted in person or online based on the preferences and requirements of the course. The session covers:
+
+* Overview of HPC concepts
+* Accessing and navigating the HPC cluster
+* Basic job submission and monitoring
+* Filesystems on HPC cluster
+* Conclusion
+
+By incorporating HPC resources into your course, you provide students with the opportunity to engage in hands-on, real-world applications of computational concepts. The HPC environment enhances the learning experience and prepares students for challenges in data-driven and computationally intensive fields. For specific requests or to schedule an HPC introduction session, please contact the [HPC Facilitator](contact.md).
+
+
+
+
+
+