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: articles/lab-services/how-to-setup-lab-gpu.md
+14-12Lines changed: 14 additions & 12 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -7,7 +7,7 @@ ms.service: azure-lab-services
7
7
author: RoseHJM
8
8
ms.author: rosemalcolm
9
9
ms.topic: how-to
10
-
ms.date: 04/24/2023
10
+
ms.date: 11/13/2024
11
11
---
12
12
13
13
# Set up a lab with GPU virtual machines in Azure Lab Services
@@ -20,7 +20,7 @@ In this article, you learn how to choose between the different GPU-based virtual
20
20
21
21
## Choose between visualization and compute GPU sizes
22
22
23
-
When you create a lab in Azure Lab Services, you have to select a virtual machine size. Choose the right virtual machine size, based on the usage scenario or [class type](./class-types.md).
23
+
When you create a lab in Azure Lab Services, you have to select a virtual machine size. Choose the right virtual machine size, based on the usage scenario.
24
24
25
25
:::image type="content" source="./media/how-to-setup-gpu/lab-gpu-selection.png" alt-text="Screenshot of the New lab window for creating a new lab in the Lab Services website, highlighting the VM sizes dropdown.":::
26
26
@@ -36,29 +36,31 @@ Azure Lab Services has two GPU-based virtual machines size categories:
36
36
37
37
The *compute* GPU size is intended for compute-intensive applications. For example, the [Deep Learning in Natural Language Processing class type](./class-type-deep-learning-natural-language-processing.md) uses the **Small GPU (Compute)** size. The compute GPU is suitable for this type of class, because lab users apply deep learning frameworks and tools that are provided by the [Data Science Virtual Machine image](https://azuremarketplace.microsoft.com/en-us/marketplace/apps?search=Data%20science%20Virtual%20machine&page=1&filters=microsoft%3Blinux) to train deep learning models with large sets of data.
38
38
39
-
| Size | vCPUs | RAM | Description |
40
-
| ---- | ----- | --- | ----------- |
41
-
| Small GPU (Compute) | 6 vCPUs | 112 GB RAM |[Standard_NC6s_v3](/azure/virtual-machines/ncv3-series). This size supports both Windows and Linux and is best suited for compute-intensive applications such as artificial intelligence (AI) and deep learning. |
39
+
| Size | vCPUs | Memory (GB) | Series | Suggested use | GPU/Accelerator | Accelerator Memory (GB) |
40
+
| - | - | - | - | - | - | - |
41
+
| Small GPU (Compute) | 8 | 56 |[NC8as_T4_v3](/azure/virtual-machines/nct4-v3-series)| AI & deep learning | NVIDIA Tesla T4 | 16 |
42
+
| Alternative Small GPU (Compute) | 6 | 112 |[NC6s_v3](/azure/virtual-machines/ncv3-series)| AI & deep learning | NVIDIA Tesla V100 | 16 |
42
43
43
44
### Visualization GPU sizes
44
45
45
46
The *visualization* GPU sizes are intended for graphics-intensive applications. For example, the [SOLIDWORKS engineering class type](./class-type-solidworks.md) shows using the **Small GPU (Visualization)** size. The visualization GPU is suitable for this type of class, because lab users interact with the SOLIDWORKS 3D computer-aided design (CAD) environment for modeling and visualizing solid objects.
46
47
47
-
| Size | vCPUs | RAM | Description |
48
-
| ---- | ----- | --- | ----------- |
49
-
| Small GPU (Visualization) | 8 vCPUs | 28 GB RAM |[Standard_NV8as_v4](/azure/virtual-machines/nvv4-series). This size is best suited for remote visualization, streaming, gaming, and encoding that use frameworks such as OpenGL and DirectX. Currently, this size supports Windows only. |
50
-
| Medium GPU (Visualization) | 12 vCPUs | 112 GB RAM |[Standard_NV12s_v3](/azure/virtual-machines/nvv3-series). This size supports both Windows and Linux. It's best suited for remote visualization, streaming, gaming, and encoding that use frameworks such as OpenGL and DirectX. |
48
+
| Size | vCPUs | Memory (GB) | Series | Suggested use | GPU/Accelerator | Accelerator Memory (GB) |
## Ensure that the appropriate GPU drivers are installed
53
57
54
58
To take advantage of the GPU capabilities of your lab VMs, ensure that the appropriate GPU drivers are installed. In the lab creation wizard, when you select a GPU VM size, you can select the **Install GPU drivers** option. This option is enabled by default.
55
59
56
60
:::image type="content" source="./media/how-to-setup-gpu/lab-gpu-drivers.png" alt-text="Screenshot of the New lab page in the Lab Services website, highlighting the Install GPU drivers option.":::
57
61
58
-
When you select **Install GPU drivers**, it ensures that recently released drivers are installed for the type of GPU and image that you selected.
62
+
When you select **Install GPU drivers**, it ensures that recently released drivers are installed for the type of GPU and image that you selected.
59
63
60
-
- When you select the Small GPU *(Compute)* size, your lab VMs are powered by the [NVIDIA Tesla V100 GPU](https://www.nvidia.com/en-us/data-center/v100/) GPU. In this case, recent Compute Unified Device Architecture (CUDA) drivers are installed, which enables high-performance computing.
61
-
- When you select the Small GPU *(Visualization)* size, your lab VMs are powered by the AMD Radeon Instinct MI25 Accelerator GPU. In this case, recent AMD GPU drivers are installed, which enables the use of graphics-intensive applications.
62
64
- When you select the Medium GPU *(Visualization)* size, your lab VMs are powered by the [NVIDIA Tesla M60](https://images.nvidia.com/content/tesla/pdf/188417-Tesla-M60-DS-A4-fnl-Web.pdf) GPU and [GRID technology](https://www.nvidia.com/content/dam/en-zz/Solutions/design-visualization/solutions/resources/documents1/NVIDIA_GRID_vPC_Solution_Overview.pdf). In this case, recent GRID drivers are installed, which enables the use of graphics-intensive applications.
0 commit comments