fix: Unpublish HowTo guide on vGPU-powered servers

docs/howto/openstack/nova/.pages

@@ -1,7 +1,6 @@
 title: Nova (compute service)
 nav:
   - new-server.md
-  - new-vgpu-server.md
   - server-group.md
   - config-drive.md
   - resize-server.md

docs/howto/openstack/nova/new-server.md

@@ -44,9 +44,8 @@ file first](../../getting-started/enable-openstack-cli.md).
     _server profile_. Currently, four profiles are available:
 
     *  _Generic_,
-    * _High CPU_,
-    * _Low Latency Disk_, and
-    *  _GPU_.
+    * _High CPU_, and
+    * _Low Latency Disk_.
 
     Selecting one of those gives you access to a subset of all available
     [_flavors_](../../../reference/flavors/index.md).
@@ -60,8 +59,6 @@ file first](../../getting-started/enable-openstack-cli.md).
       applications.
     * The _Low Latency Disk_ profile is for servers that boot off of a local
       low-latency disk.
-    * The _GPU_ profile is for servers that boot off of a local SSD disk
-      and, most importantly, have direct access to an NVIDIA vGPU.
 
     We should point out that not _all_ profiles may be available in any
     region. Choose the _Generic_ profile for now, which is available in any
@@ -82,7 +79,7 @@ file first](../../getting-started/enable-openstack-cli.md).
     on the server profile you have already selected, the boot target will be
     either _Volume_ or _Ephemeral_. More specifically...
 
-    * for the _Generic_, _High CPU_, and _GPU_ profiles the boot target is
+    * for the _Generic_ and _High CPU_ profiles the boot target is
       _Volume_, and
     * for the _Low Latency Disk_ profile the boot target is _Ephemeral_.
 

docs/reference/features/compliant.md

@@ -10,13 +10,6 @@
 > :material-close: Feature is not available
 
 
-## Virtualization
-
-|                                               | Sto1HS                | Sto2HS                |
-| -------------                                 | ----------------      | --------------------- |
-| [Virtual GPU](../flavors/index.md#compute-tiers)   | :material-timer-sand: | :material-timer-sand: |
-
-
 ## Block storage
 
 |                                                                 | Sto1HS           | Sto2HS           |

docs/reference/features/public.md

@@ -18,7 +18,6 @@
 ## Virtualization
 |                                                              | Kna1                  | Sto2                  | Fra1             | Dx1              | Tky1             |
 | -------------                                                | ----------------      | --------------------- | ---------------- | ---------------- | ---------------- |
-| [Virtual GPU](../../howto/openstack/nova/new-vgpu-server.md) | :material-timer-sand: | :material-timer-sand: | :material-close: | :material-close: | :material-close: |
 | [Physical CPUs](../flavors/index.md#compute-tiers)           | :material-close:      | :material-timer-sand: | :material-close: | :material-close: | :material-close: |
 
 

docs/reference/flavors/index.md

@@ -45,9 +45,6 @@ general-purpose compute instance with 4 cores and 32 GiB RAM.
   latency for I/O intensive applications, but instances launched with
   these flavors must configure their own high availability and data
   replication.
-* `g`: Virtual GPU. Instances launched with matching flavors have
-  access to a
-  [GPU](https://en.wikipedia.org/wiki/Graphics_processing_unit).
 * `c`: Physical CPUs. Instances launched with matching flavors will be
   assigned physical CPU cores, rather then virtual ones. These flavors
   are recommended for CPU-intensive workloads, or when there is a

docs/reference/images/index.md

@@ -17,16 +17,8 @@ Image names in {{brand}} follow a convention, which can be summarized as `${NAME
 * `NAME`: Operating system name, such as `Ubuntu`, `Debian`, `Rocky`, etc.
 * `VERSION_ID`: Operating system version, as in `22.04`, `11`, `9`, etc.
 * `CODENAME`: Operating system codename, if present, like `Jammy Jellyfish`, `Bullseye`, etc.
-   Modified codenames may reflect that some images serve special purposes, such as vGPU support.
-   See below for details.
 * `ARCH`: Platform architecture for which the operating system was built, for example: `x86_64`, `aarch64`, etc.
 
-### Special codenames
-
-* `NVGRID`: Images with `NVGRID` in the codename include pre-installed Nvidia GRID drivers and license files.
-  These images allow you to use [CUDA](https://en.wikipedia.org/wiki/CUDA), and enable other computational features of vGPUs.
-  You should use these images together with [flavors](../flavors/index.md) using the `g` [compute tier](../flavors/index.md#Compute_tiers).
-
 ## Tags and properties
 
 Each public image is assigned a specific set of tags and properties.