Skip to content

Commit e87ee87

Browse files
Merge pull request #15513 from sethmanheim/ash-gpu
Hub: specify AMD version
2 parents da91f0f + 8d1b519 commit e87ee87

File tree

1 file changed

+19
-19
lines changed

1 file changed

+19
-19
lines changed

azure-stack/user/gpu-vms-about.md

Lines changed: 19 additions & 19 deletions
Original file line numberDiff line numberDiff line change
@@ -5,7 +5,7 @@ author: sethmanheim
55
ms.author: sethm
66
ms.service: azure-stack
77
ms.topic: reference
8-
ms.date: 06/03/2024
8+
ms.date: 07/18/2024
99
ms.reviewer: unknown
1010
ms.lastreviewed: 4/28/2021
1111

@@ -15,22 +15,22 @@ ms.lastreviewed: 4/28/2021
1515

1616
# Graphics processing unit (GPU) virtual machine (VM) on Azure Stack Hub
1717

18-
This article describes which graphics processing unit (GPU) models are supported on an Azure Stack Hub integrated system. You can also find instructions on installing the drivers used with the GPUs. GPU support in Azure Stack Hub enables solutions such as artificial intelligence, training, inference, and data visualization. The AMD Radeon Instinct MI25 can be used to support graphic-intensive applications such as Autodesk AutoCAD.
18+
This article describes which graphics processing unit (GPU) models are supported on an Azure Stack Hub integrated system. The article also contains instructions on installing the drivers used with the GPUs. GPU support in Azure Stack Hub enables solutions such as artificial intelligence, training, inference, and data visualization. The AMD Radeon Instinct MI25 can be used to support graphic-intensive applications such as Autodesk AutoCAD.
1919

20-
You can choose from three GPU models. They are available in NVIDIA V100, NVIDIA T4 and AMD MI25 GPUs. These physical GPUs align with the following Azure N-Series virtual machine (VM) types as follows:
20+
You can choose from three GPU models. They are available in NVIDIA V100, NVIDIA T4, and AMD MI25 GPUs. These physical GPUs align with the following Azure N-Series virtual machine (VM) types as follows:
2121

2222
- [NCv3](/azure/virtual-machines/ncv3-series)
2323
- [NVv4 (AMD MI25)](/azure/virtual-machines/nvv4-series)
2424
- [NCasT4_v3](/azure/virtual-machines/nct4-v3-series)
2525

2626
::: moniker range="<=azs-2002"
2727
> [!WARNING]
28-
> GPU VMs are not supported in this release. You will need to upgrade to Azure Stack Hub 2005 or later. In addition, your Azure Stack Hub hardware must have physical GPUs.
28+
> GPU VMs are not supported in this release. You must upgrade to Azure Stack Hub 2005 or later. In addition, your Azure Stack Hub hardware must have physical GPUs.
2929
::: moniker-end
3030

3131
## NCv3
3232

33-
NCv3-series VMs are powered by NVIDIA Tesla V100 GPUs. Customers can take advantage of these updated GPUs for traditional HPC workloads such as reservoir modeling, DNA sequencing, protein analysis, Monte Carlo simulations, and others.
33+
NCv3-series VMs are powered by NVIDIA Tesla V100 GPUs. Customers can take advantage of these updated GPUs for traditional HPC workloads such as reservoir modeling, DNA sequencing, protein analysis, Monte Carlo simulations, and others.
3434

3535
| Size | vCPU | Memory: GiB | Temp storage (SSD) GiB | GPU | GPU memory: GiB | Max data disks | Max NICs |
3636
|---|---|---|---|---|---|---|---|
@@ -40,7 +40,7 @@ NCv3-series VMs are powered by NVIDIA Tesla V100 GPUs. Customers can take advant
4040

4141
## NVv4
4242

43-
The NVv4-series virtual machines are powered by AMD Radeon Instinct MI25 GPUs. With NVv4-series Azure Stack Hub is introducing virtual machines with partial GPUs. This size can be used for GPU accelerated graphics applications and virtual desktops. NVv4 virtual machines currently support only Windows guest operating system.
43+
The NVv4-series virtual machines are powered by AMD Radeon Instinct MI25 GPUs. With the NVv4-series, Azure Stack Hub introduces virtual machines with partial GPUs. This size can be used for GPU accelerated graphics applications and virtual desktops. NVv4 virtual machines currently support only the Windows guest operating system.
4444

4545
| Size | vCPU | Memory: GiB | Temp storage (SSD) GiB | GPU | GPU memory: GiB | Max data disks | Max NICs |
4646
| --- | --- | --- | --- | --- | --- | --- | --- |
@@ -77,20 +77,20 @@ The NC_A100 series VMs are powered by NVIDIA Ampere A100 GPUs, the successor of
7777

7878
## Capacity planning
7979

80-
The Azure Stack Hub capacity planner has been updated to support GPU configurations. It is accessible on https://aka.ms/azstackcapacityplanner.
80+
The Azure Stack Hub capacity planner was updated to support GPU configurations. [It's accessible here](https://aka.ms/azstackcapacityplanner).
8181

8282
## Adding GPUs on an existing Azure Stack Hub
8383

84-
Azure Stack Hub now supports adding GPUs to any existing system. To do this, execute stop-azurestack, run through the procedure of stop-azurestack, add GPUs, and then run **start-azurestack** until completion. If the system already had GPUs, then any previously created GPU VMs will need to be **stop-deallocated** and then **restarted**.
84+
Azure Stack Hub now supports adding GPUs to any existing system. To add a GPU, run `stop-azurestack`, run through the procedure of `stop-azurestack`, add GPUs, and then run `start-azurestack` until completion. If the system already had GPUs, then any previously created GPU VMs must be `stop-deallocated` and then restarted.
8585

86-
## Patch and update, FRU behavior of VMs
86+
## Patch and update, FRU behavior of VMs
8787

88-
GPU VMs will undergo downtime during operations such as patch and update (PnU) and hardware replacement (FRU) of Azure Stack Hub. The following table covers the state of the VM as observed during these activities and the manual action you can do to make these VMs available after the operation.
88+
GPU VMs undergo downtime during operations such as patch and update (PnU) and hardware replacement (FRU) of Azure Stack Hub. The following table covers the state of the VM as observed during these activities and the manual action you can do to make these VMs available after the operation.
8989

9090
| Operation | PnU - Full Update, OEM update | FRU |
9191
| --- | --- | --- |
9292
| VM state | Unavailable during update. Can be made available with manual operation. VM is automatically online post update. | Unavailable during FRU. Can be made available with manual operation. VM needs to be brought back up after FRU|
93-
| Manual operation | If the VM needs to be made available during the update, if there are available GPU partitions, the VM can be restarted from the portal by clicking the **Restart** button. VM will automatically come back up post update | VM is not available during FRU. If there are available GPUs, VM may be stop-deallocated and restarted during FRU. Post FRU completion, VM needs to be stop-deallocated using the **Stop** button and started back up using the **Start** button.|
93+
| Manual operation | If the VM needs to be made available during the update, if there are available GPU partitions, the VM can be restarted from the portal by clicking the **Restart** button. VM automatically comes back up post update. | VM is not available during FRU. If there are available GPUs, VM may be stop-deallocated and restarted during FRU. Post FRU completion, the VM must be `stop-deallocated` using the **Stop** button, then restarted using the **Start** button.|
9494

9595
## Guest driver installation
9696

@@ -116,15 +116,15 @@ Set-AzureRmVMExtension -Location $Location `
116116
-Verbose
117117
```
118118

119-
Depending on the OS, type and connectivity of your Azure Stack Hub GPU VM, you will need to modify with the settings below.
119+
Depending on the OS, type and connectivity of your Azure Stack Hub GPU VM, you must replace these values with the settings below.
120120

121121
### AMD MI25
122122

123123
The guest driver version must match the Azure Stack Hub version, regardless of the connectivity state. Using newer versions not aligned with the Azure Stack Hub version can cause usability issues.
124124

125125
| Azure Stack Hub Version | AMD Guest driver |
126126
|--------------------------------|-------------------------|
127-
| 2206 | [21.Q2-1](https://download.microsoft.com/download/4/e/a/4ea28d3f-28e2-4eaa-8ef2-4f7d32882a0b/AMD-Azure-NVv4-Driver-21Q2-1.exe), [20.Q4-1](https://download.microsoft.com/download/0/e/6/0e611412-093f-40b8-8bf9-794a1623b2be/AMD-Azure-NVv4-Driver-20Q4-1.exe) |
127+
| 2206 and later | [21.Q2-1](https://download.microsoft.com/download/4/e/a/4ea28d3f-28e2-4eaa-8ef2-4f7d32882a0b/AMD-Azure-NVv4-Driver-21Q2-1.exe), [20.Q4-1](https://download.microsoft.com/download/0/e/6/0e611412-093f-40b8-8bf9-794a1623b2be/AMD-Azure-NVv4-Driver-20Q4-1.exe) |
128128
| 2108 | [21.Q2-1](https://download.microsoft.com/download/4/e/a/4ea28d3f-28e2-4eaa-8ef2-4f7d32882a0b/AMD-Azure-NVv4-Driver-21Q2-1.exe), [20.Q4-1](https://download.microsoft.com/download/0/e/6/0e611412-093f-40b8-8bf9-794a1623b2be/AMD-Azure-NVv4-Driver-20Q4-1.exe) |
129129
| 2102 | [21.Q2-1](https://download.microsoft.com/download/4/e/a/4ea28d3f-28e2-4eaa-8ef2-4f7d32882a0b/AMD-Azure-NVv4-Driver-21Q2-1.exe), [20.Q4-1](https://download.microsoft.com/download/0/e/6/0e611412-093f-40b8-8bf9-794a1623b2be/AMD-Azure-NVv4-Driver-20Q4-1.exe) |
130130

@@ -134,7 +134,7 @@ Use the PowerShell script in the previous section with the appropriate driver ty
134134

135135
#### Disconnected
136136

137-
Since the extension pulls the driver from a location on the internet, a VM that is disconnected from the external network cannot access it. You can [download the driver from the previous table](#amd-mi25) and upload to a storage account in your local network that's accessible to the VM.
137+
Since the extension pulls the driver from a location on the internet, a VM that is disconnected from the external network can't access it. You can [download the driver from the previous table](#amd-mi25) and upload to a storage account in your local network that's accessible to the VM.
138138

139139
Add the AMD driver to a storage account and specify the URL to that account in `Settings`. These settings must be used in the **Set-AzureRMVMExtension** cmdlet. For example:
140140

@@ -161,7 +161,7 @@ $Settings = @{
161161

162162
### Use case: compute/CUDA - Connected
163163

164-
CUDA drivers do not need a license server and do not need modified settings.
164+
CUDA drivers don't need a license server and don't need modified settings.
165165

166166
### Use case: compute/CUDA - Disconnected
167167

@@ -180,16 +180,16 @@ $Settings = @{
180180

181181
**Linux:**
182182

183-
You will need to reference some URLs for your settings.
183+
You must reference some URLs for your settings:
184184

185185
| URL | Notes |
186186
| --- | --- |
187-
| PUBKEY_URL | The PUBKEY_URL is the public key for the Nvidia driver repository not for the Linux VM. It is used to install driver for Ubuntu. |
188-
| DRIVER_URL | DRIVER_URL is the URL to download the Nvidia driver's repository information and it is added to the Linux VM's list of repos. |
187+
| PUBKEY_URL | The PUBKEY_URL is the public key for the Nvidia driver repository not for the Linux VM. It's used to install driver for Ubuntu. |
188+
| DRIVER_URL | DRIVER_URL is the URL to download the Nvidia driver's repository information and is added to the Linux VM's list of repos. |
189189

190190
Add the URLs to your settings.
191191

192-
```powershell
192+
```powershell
193193
$Settings=@{
194194
"isCustomInstall"=$true;
195195
"DRIVER_URL"="https://go.microsoft.com/fwlink/?linkid=874273";

0 commit comments

Comments
 (0)