File: content/manuals/engine/containers/gpu.md
Issue
The page contains a vague note about GPU access restrictions:
Note
NVIDIA GPUs can only be accessed by systems running a single engine.
This statement is unclear and would confuse readers because:
"Single engine" is not defined - does it mean:
Only one Docker Engine instance per host?
Not supported in Docker Swarm (multiple engines)?
Only one container can access the GPU at a time?
Something else entirely?
The note appears without context or explanation
There's no link to documentation explaining this limitation or its implications
Why this matters
Readers trying to use GPUs with Docker will encounter this note and won't know:
Whether they can run multiple containers with GPU access on the same host
Whether this is a Docker Swarm limitation
What happens if they violate this constraint
How to configure their system to comply with this requirement
The lack of clarity could lead readers to make incorrect assumptions about GPU support in their Docker deployments.
Suggested fix
Clarify what "single engine" means with specific language, such as:
If it means Swarm: "NVIDIA GPUs are not supported in Docker Swarm mode. Use standalone Docker Engine for GPU access."
If it means one engine per host: "Only one Docker Engine instance per host can access NVIDIA GPUs."
If it means something else: Provide the specific constraint with an explanation
Also consider adding a link to relevant documentation about GPU limitations or multi-engine scenarios.
Found by nightly documentation quality scanner
File:
content/manuals/engine/containers/gpu.mdIssue
The page contains a vague note about GPU access restrictions:
Note
NVIDIA GPUs can only be accessed by systems running a single engine.
This statement is unclear and would confuse readers because:
"Single engine" is not defined - does it mean:
The note appears without context or explanation
There's no link to documentation explaining this limitation or its implications
Why this matters
Readers trying to use GPUs with Docker will encounter this note and won't know:
The lack of clarity could lead readers to make incorrect assumptions about GPU support in their Docker deployments.
Suggested fix
Clarify what "single engine" means with specific language, such as:
Also consider adding a link to relevant documentation about GPU limitations or multi-engine scenarios.
Found by nightly documentation quality scanner