From a3eadbba5407cc686586691d4339835e4ddcf911 Mon Sep 17 00:00:00 2001
From: Milson Munakami <milsonmun@yahoo.com>
Date: Mon, 22 Jul 2024 17:30:05 -0400
Subject: [PATCH] added example for the quota resources required for a A100SXM4
 flavor

---
 .../allocation/allocation-change-request.md   | 46 +++++++++++++++++--
 1 file changed, 43 insertions(+), 3 deletions(-)

diff --git a/docs/get-started/allocation/allocation-change-request.md b/docs/get-started/allocation/allocation-change-request.md
index dbe17934..4e313a49 100644
--- a/docs/get-started/allocation/allocation-change-request.md
+++ b/docs/get-started/allocation/allocation-change-request.md
@@ -31,7 +31,15 @@ your resource allocation for the selected project.
 !!! info "Information"
     PI or project managers can put the new values on the textboxes for **ONLY**
     quota attributes they want to change others they can be left **blank** so those
-    quotas will not get changed!
+    quotas will not get changed! To use GPU resources on your VM, you need to
+    specify the number of GPUs in the "OpenStack GPU Quota" attribute. Additionally,
+    ensure that your other quota attributes, namely "OpenStack Compute vCPU Quota"
+    and "OpenStack Compute RAM Quota (MiB)" have sufficient resources to meet the
+    **vCPU** and **RAM** requirements for one of the GPU tier-based flavors.
+
+    Refer to the [GPU Tier documentation](../../openstack/create-and-connect-to-the-VM/flavors.md#3-gpu-tier)
+    for specific requirements and further details on the flavors available for GPU
+    usage.
 
 ### Allocation Change Requests for OpenStack Project
 
@@ -63,6 +71,36 @@ ensure you know which GPU-based flavor you want to use, then submit an
 [allocation change request](#request-change-resource-allocation-attributes-for-openstack-project)
 to adjust your current allocation to fit the flavor's resource requirements.
 
+!!! tip "Resource Requirements for Launching a VM with "NVIDIA A100 SXM4 40GB" Flavor."
+    Based on the [GPU Tier documentation](../../openstack/create-and-connect-to-the-VM/flavors.md#i-nvidia-a100-sxm4-40gb),
+    NERC provides two variations of NVIDIA A100 SXM4 40GB flavors:
+
+    1. **`gpu-su-a100sxm4.1`**: Includes 1 NVIDIA A100 GPU
+    2. **`gpu-su-a100sxm4.2`**: Includes 2 NVIDIA A100 GPUs
+
+    You should select the flavor that best fits your resource needs and ensure your
+    OpenStack quotas are appropriately configured for the chosen flavor. To use
+    a GPU-based VM flavor, choose the one that best fits your resource needs and
+    make sure your OpenStack quotas meet the required specifications:
+
+    - For the **`gpu-su-a100sxm4.1`** flavor:
+        - **vCPU**: 32
+        - **RAM (GiB)**: 240
+
+    - For the **`gpu-su-a100sxm4.2`** flavor:
+        - **vCPU**: 64
+        - **RAM (GiB)**: 480
+
+    Ensure that your OpenStack resource quotas are configured as follows:
+
+    - **OpenStack GPU Quota**: Meets or exceeds the number of GPUs required by the
+    chosen flavor.
+    - **OpenStack Compute vCPU Quota**: Meets or exceeds the vCPU requirement.
+    - **OpenStack Compute RAM Quota (MiB)**: Meets or exceeds the RAM requirement.
+
+    Properly configure these quotas to successfully launch a VM with the selected
+    "gpu-su-a100sxm4" flavor.
+
 - We recommend using [ubuntu-22.04-x86_64](../../openstack/create-and-connect-to-the-VM/images.md#nerc-images-list)
 as the image for your GPU-based instance because we have tested the NVIDIA driver
 with this image and obtained good results. That said, it is possible to run a
@@ -86,10 +124,12 @@ Then submitting the change request, this will notify the NERC admin about it. Pl
 wait untill the NERC admin approves/ deny the change request to see the change on
 your resource allocation for the selected project.
 
-!!! info "Information"
+!!! tip "Important Information"
     PI or project managers can put the new values on the textboxes for **ONLY**
     quota attributes they want to change others they can be left **blank** so those
-    quotas will not get changed!
+    quotas will not get changed! In order to use GPU resources on your pod, you
+    must specify the number of GPUs you want to use in the "OpenShift Request on
+    GPU Quota" attribute.
 
 ### Allocation Change Requests for OpenShift Project