Merge pull request #2929 from JdeRobot/dph/windows_gpu_dev_docs

Update GPU acceleration docs for devs

docs/gpu_acceleration.md

@@ -8,23 +8,30 @@ ROS and Gazebo can be accelerated within RoboticsAcademy thanks to VirtualGL if
 ## Linux
 
 ### Auto
-If the PC has several GPUs, it will select them in order: NVIDIA, Intel, Only CPU. 
+If the PC has several GPUs, it will select them in order: NVIDIA, Intel, Only CPU.
 ```
-docker run --rm -it $(nvidia-smi >/dev/null 2>&1 && echo "--gpus all" || echo "") --device /dev/dri -p 7164:7164 -p 6080:6080 -p 1108:1108 -p 7163:7163 jderobot/robotics-backend
+docker run --rm -it $(nvidia-smi >/dev/null 2>&1 && echo "--gpus all" || echo "") --device /dev/dri -p 7164:7164 -p 6080:6080 -p 1108:1108 -p 7163:7163 --link academy_db jderobot/robotics-academy:latest
 ```
 
 
 ### Intel
 For Linux machines and Intel GPUs, acceleration can be achieved by simply setting the ```--device``` argument when running the Docker container:
 ```
-docker run --rm -it --device /dev/dri -p 7164:7164 -p 6080:6080 -p 1108:1108 -p 7163:7163 jderobot/robotics-backend
+docker run --rm -it --device /dev/dri -p 7164:7164 -p 6080:6080 -p 1108:1108 -p 7163:7163 --link academy_db jderobot/robotics-academy:latest
 ```
 
 ### NVIDIA
 For NVIDIA GPUs, acceleration can be achieved by [installing the nvidia-container-runtime package](https://docs.docker.com/config/containers/resource_constraints/#gpu), and then running the command above, but adding the ```--gpus all``` flag:
 ```
-docker run --rm -it --gpus all --device /dev/dri -p 7164:7164 -p 6080:6080 -p 1108:1108 -p 7163:7163 jderobot/robotics-backend
+docker run --rm -it --gpus all --device /dev/dri -p 7164:7164 -p 6080:6080 -p 1108:1108 -p 7163:7163 --link academy_db jderobot/robotics-academy:latest
 ```
 
 ## Windows
-Pending validation.
+For Windows machines, acceleration can be achieved for NVIDIA GPUs if a valid CUDA installation is available. Useful docs for proper installation of WSL2 + CUDA + Docker Desktop:
+- [WSL2 + CUDA](https://learn.microsoft.com/en-us/windows/ai/directml/gpu-cuda-in-wsl)
+- [WSL2 + Docker Desktop](https://docs.docker.com/desktop/features/wsl/)
+
+Once these requirements are ready, you should be able to run Robotics Academy with GPU acceleration as follows:
+```
+docker run --rm -it --gpus all -v /usr/lib/wsl:/usr/lib/wsl -e LD_LIBRARY_PATH=/usr/lib/wsl/lib --device /dev/dri -p 6080:6080 -p 1108:1108 -p 7163:7163 -p 7164:7164 --link academy_db jderobot/robotics-academy:latest
+```